Blog|Support|Login
SEOScanPRO

Register FreeRegister

Author: SEOScanPRO

  • AI agents run blind inside most enterprise networks, and firewalls built for the old web cannot follow them

    AI agents run blind inside most enterprise networks, and firewalls built for the old web cannot follow them

    Nearly half of organizations have zero visibility into the machine-to-machine traffic their AI agents generate, and the firewalls and API gateways most enterprises rely on were not built to look inside a prompt. The result is a specific gap in security monitoring, one that vendors are now trying to close from inside the infrastructure companies already run.

    That gap matters more as agents take on tasks such as purchasing access to data and online services, handling customer interactions, and completing work without a person approving every step. A log showing that one service contacted another cannot, on its own, explain whether the agent followed the company’s instructions.

    What the new blind spot actually looks like

    One recent study found that 48.9% of organizations have no way to monitor what their autonomous agents are doing across connected systems. Legacy web application firewalls and standard API gateways were built around attack signatures, rate limits, and predictable human sessions. An agent can improvise a new sequence of otherwise legitimate requests without matching a known attack signature, so those tools have nothing to flag.

    An approved API call is also not automatically an approved business decision. A support agent that uses valid credentials to pull a customer file and then includes it in a reply to someone who should not receive it performs no action that resembles a conventional intrusion. Encryption adds a second obstacle, since AI traffic over HTTPS needs the right certificates and policies before a firewall can decrypt and inspect it at all.

    Even after decryption, a readable prompt is not the same as an understood prompt. Decryption exposes the text, but it does not show whether the instructions are safe.

    Two different things called an AI firewall

    The term covers two distinct products. An AI-powered firewall uses machine learning to detect conventional network threats. A firewall built to protect AI inspects prompts and agent interactions for AI-specific harm, including prompt injection, a technique that can turn a document or webpage into instructions an agent follows, and that has been used in recent attacks against coding agents.

    How Check Point is approaching the gap

    Check Point’s AI Network Firewall, announced this past July, adds AI-specific inspection to existing firewall infrastructure. The product discovers and classifies employee use of generative-AI tools, AI agent activity, Model Context Protocol traffic, and traffic to and from AI applications, then applies real-time inspection to that activity. Model Context Protocol is the industry standard for connecting agents to tools and data.

    The product aims to identify sensitive data heading toward a public AI tool and to flag a manipulated prompt attempting to trigger unintended behavior. Check Point’s broader AI security stack incorporates technology from Lakera, the AI security startup Check Point acquired in 2025, which supplies runtime protection against prompt attacks.

    What stands out is the deployment model. Customers can use their existing firewall infrastructure without adding new hardware or software, bringing AI controls into their established management environment. Security teams can start governing AI traffic without first deploying and maintaining a separate system.

    How Nightfall AI is approaching the same problem

    Nightfall AI’s Firewall for AI takes a different route. The company describes its standalone offering as a client wrapper around generative-AI interactions, using APIs and software development kits to inspect content before it reaches a model. It scans for personally identifiable information, payment-card details, health information, and secrets, so sensitive material can be removed before an application forwards a prompt. Nightfall also offers prompt-injection protection and conversational guardrails separately, checks that cover conversation content and signals such as model-response refusals.

    Finding a payment card number and recognizing an attempt to redirect a model are different security tasks, and the two vendors address them in different places on the network.

    What this means for organizations running agents

    Check Point’s argument is that AI-specific protection belongs inside infrastructure a company already runs. Nightfall’s argument is that AI interactions warrant a dedicated layer inside application workflows. Neither placement, on its own, guarantees that every relevant interaction will be inspected.

    For companies that need to protect their AI systems, the practical questions concern coverage, intervention, and policy enforcement, not which product approach seems freshest. Both point to a wider focus on reliable AI infrastructure rather than model performance alone. An integrated control may fit established operations, while an application-level wrapper gives developers a specific point at which to filter model-bound data.

    A business that cannot observe its agents’ interactions cannot confidently assess whether those agents are staying within their remit. Whichever architecture gains ground, the meaningful advance will be tooling that connects an instruction to an action and applies policy before harm occurs. What matters is whether security tools can see what AI systems are actually doing, and govern it, rather than logging the traffic after the fact.

    FAQ

    What percentage of organizations cannot see what their AI agents are doing?

    48.9% of organizations have zero visibility into the machine-to-machine traffic their AI agents generate, according to a recent study cited on the gap.

    Why can’t traditional firewalls monitor AI agent traffic?

    Legacy web application firewalls and standard API gateways were built around attack signatures, rate limits, and predictable human sessions. An agent can improvise a new sequence of legitimate requests without matching a known attack signature, and HTTPS encryption requires the right certificates and policies before any inspection can happen at all.

    What is the difference between an AI-powered firewall and a firewall built to protect AI?

    An AI-powered firewall uses machine learning to detect conventional network threats. A firewall built to protect AI inspects prompts and agent interactions for AI-specific harm, including prompt injection, where a document or webpage is turned into instructions an agent follows.

    This article summarizes reporting from thenextweb.com.

  • Why Is SEO Important? 11 Reasons It Still Matters in 2026

    Why Is SEO Important? 11 Reasons It Still Matters in 2026

    SEO remains one of the highest-return investments in digital marketing because rankings compound: the content optimized today keeps earning clicks months and years later, while AI search grows on top of the same foundation. This guide covers what SEO is, how it works, and 11 concrete reasons it still matters in a search landscape that now includes AI Overviews, AI Mode, and ChatGPT.

    What Is SEO?

    Search engine optimization is the process of improving a website’s content, structure, and overall online presence to earn greater organic, unpaid visibility in search engine results pages. Done well, SEO earns visibility in Google, including in AI Overviews and AI Mode, and in organic rankings.

    SEO is one of two main pillars of search engine marketing. The other is pay-per-click marketing, where a business pays to appear as a sponsored result.

    How Does SEO Work?

    SEO works by helping search engines understand what a page is about so it can rank for appropriate queries. Google wants to rank pages that satisfy search intent, demonstrate experience, expertise, authoritativeness, and trustworthiness (E-E-A-T), load quickly, and are easy to navigate. These same signals help AI systems decide which results to surface, so SEO work also builds visibility in AI-generated answers.

    The most common SEO tactics include keyword research, content creation, link building, and technical SEO.

    Why Is SEO Important?

    SEO matters because it improves organic visibility in the search results, which translates into more brand awareness, website traffic, and ultimately more sales. The 11 reasons below explain each benefit in detail.

    1. Search Engines Generate Massive Traffic

    Google processes roughly 16.4 billion searches per day, and organic search generated over one trillion visits in 2025, more than any other channel. AI traffic grew 66% that year but from a much smaller base, under one billion visits per month, against tens of billions from organic search. Capturing the available organic clicks in a niche is what separates winning brands from invisible ones.

    2. SEO Can Be Highly Cost Effective

    SEO is more cost-effective than paid channels because a well-ranking page can keep earning visibility for months or years without ongoing ad spend. The main costs are content creation, updates over time, and site maintenance. A digital agency helped one B2B services firm grow organic monthly traffic from roughly 4.1K to 12.4K visits between 2025 and 2026, with organic keyword visibility up more than 103%. Traffic has kept climbing since.

    3. Search Visibility Builds Brand Awareness and Authority

    Brands that frequently appear in the search results accumulate awareness simply through repeated exposure. A language-learning platform that ranks for thousands of topical keywords becomes the first name searchers think of, even when they never click. Repeated presence is what makes a brand feel authoritative and trustworthy, increasing the chance that searchers engage later.

    4. Competitors Are (Probably) Doing It

    If competitors are using SEO, they are capturing the visibility a business should be competing for. If they are not, that is a clear opening to dominate the niche. Comparing rankings against competitors across thousands of shared keywords shows exactly which terms others own and where the gaps are.

    5. SEO Complements Google Ads

    Appearing in both organic and paid positions gives a brand the best of both worlds. PPC ads usually sit at the top of the page and get strong visibility, but a meta-analysis found the top five organic results receive significantly more clicks overall than paid ads for the same queries. Owning both placements doubles the real estate on the results page.

    6. Reach Prospects Throughout the Buying Journey

    People turn to search at every stage, from early research to final purchase. A mattress brand’s size-comparison guide attracts roughly 164.6K organic visits per month, with keywords spanning informational and commercial intent. Ranking for informational, commercial, navigational, and transactional keywords captures demand from research all the way through to checkout.

    7. Optimization Efforts Improve the User Experience

    Many core SEO tactics are really website best practices in disguise. Google weights Core Web Vitals, including Largest Contentful Paint (LCP), Cumulative Layout Shift (CLS), and Interaction to Next Paint (INP), all metrics that measure how usable a page feels. Improving them lifts rankings and lifts the conversion rate at the same time.

    8. Keyword Research Lets You Monitor Market Trends

    Search volume data is a real-time feed of what audiences want. Search interest in ballet sneakers has climbed over recent months, a signal that footwear retailers can act on by stocking the style, publishing comparison content, or both. The same data points to rising demand in nearly every category.

    9. SEO Results Are Quantifiable

    Organic rankings, traffic, and conversions are all measurable, which makes it possible to gauge performance and prove ROI. A position-tracking tool with a visibility score running from 0% (ranking outside the top 100 for all tracked keywords) to 100% (holding the top organic spot for every tracked keyword) puts the trend in a single graph. Google Analytics 4 layers on organic traffic and organic conversions to close the loop.

    10. SEO Helps You Protect Your Brand

    Controlling what appears when people search for a brand is a defensive moat. Without SEO, a negative article or review can rank prominently in branded results and damage relationships with prospective customers before the relationship ever starts.

    11. SEO Is the Foundation for AI Search Visibility

    The work that earns rankings is the same work that earns citations in AI answers. The evaluator has shifted from a human clicking a link to an AI composing an answer, but the fundamentals have not: structured content, authoritative sourcing, and clear signals about what a page is actually about. Optimizing for AI answers, sometimes called answer engine optimization (AEO), shares those fundamentals with SEO. A few things work differently:

    • Direct answers matter more. AI systems tend to pull from content that answers a question plainly and early, rather than building up to the point.
    • Writing for query fan-out beats writing for a single keyword. AI systems often break a single prompt into multiple related sub-queries behind the scenes to build an answer, so a page optimized around one exact-match keyword can miss the broader set of questions the AI is actually trying to answer.
    • Category entry points (CEPs) anchor the strategy. CEPs start from the real-world situations that make someone think of a category in the first place. Using keywords and CEPs together gives AI systems and search engines more ways to recognize and cite the content.

    Tracking citations, mentions, visibility, and prompts in an AI visibility report closes the loop and shows what is actually being surfaced.

    Start Building an SEO Strategy

    Rankings compound, content keeps earning visibility long after publication, and AI search now reads off the same foundation. The brands that invest in SEO now are the ones that will own the results, both human and AI, for years to come.

    FAQ

    What is SEO and why is it important?

    SEO, or search engine optimization, is the process of improving a website’s content, structure, and online presence to earn greater organic visibility in search engine results. It is important because it drives sustainable, compounding traffic, builds brand awareness and authority, and now forms the foundation for being cited in AI-generated answers from systems like Google AI Overviews and ChatGPT.

    Is SEO still worth it in 2026?

    Yes. Organic search generated over one trillion visits in 2025, more than any other channel, and AI traffic grew 66% that year on top of that base. Well-ranked pages can keep earning clicks for months or years without ongoing ad spend, which makes SEO one of the highest-return investments in digital marketing.

    How does SEO help with AI search visibility?

    The same SEO work that earns rankings, including structured content, authoritative sourcing, and clear topical signals, also earns citations in AI answers. AI systems tend to pull from content that answers a question plainly and early, and they often break a single prompt into multiple related sub-queries. Writing for query fan-out and category entry points, not just one exact-match keyword, gives AI more ways to surface a page.

    Try the AI visibility report

    SEOScanPro, which includes the AI visibility report

    The AI visibility report runs a full technical audit of a site and shows the measured result behind every check. Open the AI visibility report.

    This article summarizes reporting from semrush.com.

  • Google AI Overview Links Now Route Users Into AI Mode Instead of Web Pages

    Google AI Overview Links Now Route Users Into AI Mode Instead of Web Pages

    Google is testing a change in AI Overviews where normal-looking links can pull searchers deeper into an AI experience rather than sending them to a publisher’s page. The change, spotted in late September 2026, keeps users inside Google’s conversational interface for a follow-up query instead of handing them off to the web result behind the link.

    The anchor links sit at the bottom of an AI Overview answer and look like standard citations. When clicked, they launch a follow-up question inside AI Mode rather than opening the underlying website. The behavior means a searcher who expected to read a source page gets another generated answer instead.

    What the AI Overview links actually do

    The links appear in the typical position where AI Overviews show source attributions and related queries. A mouse cursor hovering over them reveals the difference: these are not links to third-party domains. They are prompts that carry the conversation forward within Google’s AI Mode.

    A screen recording of the behavior shows two anchor links near the bottom of an AI Overview. Clicking either one triggers a fresh AI Mode response built around the question represented by the link text. The destination is not a web page, a publisher’s article, or a traditional search results page.

    Why the change matters for search traffic

    AI Overviews already generate a complete answer above the organic listings. Citation links within those answers have been one of the remaining paths for a searcher to leave the AI surface and visit the site that supplied the information. When those links instead open another AI conversation, the pathway from answer to publisher shortens further.

    The test fits a pattern Google has been building toward: keeping more of the search journey inside its own AI surfaces. A searcher who lands on an AI Overview, clicks a follow-up link, and receives another generated answer has taken two steps without visiting a single web page.

    What searchers and site owners should watch

    The behavior appears to be an early test rather than a global rollout. The linking pattern shows up in select AI Overview answers, with the anchor links styled identically to ordinary source links. For site owners, the practical effect is that some clicks that previously reached a domain may now flow into more AI Mode sessions instead.

    For searchers, the change makes the answer trail feel more conversational. A follow-up question is answered in place, with no tab switch and no wait for a separate page to load. The trade-off is a layer of generated answers between the reader and the underlying source material.

    Monitoring where AI Overview links actually point has become part of tracking how visible a site remains in AI-driven search. SEOScanPro’s AI visibility tracking shows whether an AI agent can read and use a site, which matters more as Google routes additional follow-up activity through AI Mode.

    What this means for the search results page

    The test continues the migration of search features into AI surfaces. AI Overviews already occupy the top of many results pages. AI Mode sits alongside traditional search as a separate, fully conversational experience. Links that bridge the two, carrying the user from an overview into a deeper AI session, blur the line between a hybrid results page and a pure chat product.

    If the test expands, publishers may see fewer referred visits from AI Overview citations while searchers see more generated follow-up answers in a single session. Google has not yet confirmed the test or stated how widely these links appear.

    FAQ

    Are Google AI Overview links now going to AI Mode instead of websites?

    Google is testing anchor links inside AI Overviews that open follow-up questions in AI Mode instead of taking users to the underlying web page. The links look like normal source links but keep the user inside Google’s AI experience.

    Where do these AI Mode links appear in AI Overviews?

    The links appear at the bottom of an AI Overview answer, in the area where source citations and related questions are typically shown. Clicking them launches a follow-up AI Mode query rather than opening a third-party domain.

    Will this change affect traffic from Google to publisher websites?

    If the test expands, some clicks that previously reached publisher sites through AI Overview citations could instead open more AI Mode answers. The behavior removes a step where a searcher would otherwise leave Google’s AI surface to visit a web page.

    Try the AI visibility report

    SEOScanPro, which includes the AI visibility report

    The AI visibility report runs a full technical audit of a site and shows the measured result behind every check. Open the AI visibility report.

    This article summarizes reporting from seroundtable.com.

  • Google’s new Gemini Windows app turns Alt + Space into an instant AI shortcut

    Google’s new Gemini Windows app turns Alt + Space into an instant AI shortcut

    Two quick keystrokes now bring Google’s Gemini assistant onto any Windows 10 or Windows 11 screen, no browser tab required. The new Gemini app installs like any other Windows program and binds itself to Alt + Space, replacing the shortcut normally used for a window’s system menu or the PowerToys Run launcher so the AI pops up over whatever work is already open.

    What the Gemini Windows app does

    Once installed, the Gemini app behaves like a permanent overlay on top of the active workflow. Pressing Alt + Space summons a chat window without forcing the user to switch away from the current document, browser, or game. Google has offered a macOS version of the same app for some time, and the Windows release brings the desktop experience in line with what Mac users have had.

    The app connects to other Google services, including Gmail and Drive, so queries can pull from real account data rather than working from the prompt alone. In eligible countries such as the United States and Australia, subscribers to Google AI Pro or Google AI Ultra can hand multi-step tasks to an agent called Gemini Spark. Spark is not currently available in the United Kingdom.

    Beyond chat, the Windows client carries over the full web feature set, including image generation through Nano Banana and video creation through Gemini Omni.

    Why the Alt + Space shortcut matters

    The shortcut is the whole reason the app feels different from visiting gemini.google.com in a browser. Reaching the AI becomes a reflex instead of a sequence: leave the current app, find the browser, click a tab, wait for the page to load, type the question. With Alt + Space, the chat window is open before the hand leaves the keyboard.

    That change has a side effect: the shortcut has to give up something. On most Windows installations, Alt + Space is the keyboard combination for the active window’s system menu and the default trigger for PowerToys Run, a power-user launcher for apps, files, calculator functions, and system commands. Installing the Gemini app reassigns the shortcut to itself, so users who rely on PowerToys Run will need to remap it if they want both tools at once.

    Where the app falls short

    The sidebar always shows the user’s location at the bottom, with no obvious setting to hide it. Anyone who shares desktop screenshots or screen recordings will have to crop the address out or close the sidebar entirely, since Google has not exposed a toggle for the field.

    There is no Gemini Live experience inside the Windows app yet. Live is the voice and camera mode available on iOS and Android that lets users speak to Gemini conversationally and point a phone camera at objects for the AI to describe. A desktop version that could see the active screen and answer questions about open windows, error messages, or design layouts is not here yet. For now, the workaround is to take a screenshot and upload it through the app, then ask questions about the image. Google has said additional native desktop capabilities will roll out over time.

    Who can use it and how to get it

    The Gemini app is available globally for Windows 10 and Windows 11 users. It can be downloaded directly from Google. After installation, the Alt + Space shortcut is registered automatically, and the app becomes the default destination for that key combination.

    Two things to keep in mind before installing:

    • The shortcut conflict with PowerToys Run and the system menu may matter to anyone who already relies on either tool.
    • The persistent location display in the sidebar is worth treating as a privacy consideration for screenshots, streams, and screen recordings.

    Both rough edges are manageable, and neither removes the core benefit: a one-keystroke path to Gemini on a Windows PC, with the same feature set as the web version and access to Google’s wider AI tools.

    FAQ

    What keyboard shortcut opens the new Gemini Windows app?

    Alt + Space opens the Gemini Windows app once it is installed. The app takes over that shortcut, which is also used by default for the active window’s system menu and for Microsoft PowerToys Run.

    Can the Gemini app for Windows access Gmail and Drive?

    Yes. The Windows app can connect to Gmail and Drive so that Gemini can pull information directly from a user’s Google account when answering questions.

    Does the Gemini Windows app have voice or camera features like on mobile?

    Not yet. The Windows app does not include the Gemini Live voice and camera mode that is available on iOS and Android. Users can take a screenshot, upload it through the app, and ask Gemini about what is on screen. Google has said more native desktop capabilities will arrive in future updates.

    This article summarizes reporting from techradar.com.

  • Semrush MCP Use Cases for SEO Research in AI Assistants

    Semrush MCP Use Cases for SEO Research in AI Assistants

    The Semrush MCP server connects Semrush data directly to AI assistants such as Claude and ChatGPT, so SEO research can be run in plain language. Built on the Model Context Protocol, an open standard from Anthropic that gives AI models a universal way to connect to external data sources, files, and tools, the server turns Semrush metrics into answers an assistant can return inside a conversation. This guide groups sixteen Semrush MCP use cases by workflow: keyword strategy, competitive intelligence, content optimization, and diagnostics, with copy-paste prompts ready to adapt.

    How do you set up the Semrush MCP?

    Setup takes four steps. First, check the plan: MCP access comes with Semrush One Starter, Semrush One Pro+, SEO Classic Pro, and SEO Classic Guru, each including 50,000 API units. Traffic Market reports need a separate Trends API subscription, which matters when the Traffic Analytics prompts come up.

    Second, connect from inside the AI client. In Claude, go to Settings, then Connectors, click Add, then Browse Connectors, search for Semrush MCP, and approve the permissions. In ChatGPT, go to Settings, then Apps, find Semrush, and click Connect. Both clients use OAuth, so there is no key to paste.

    Third, use the endpoint for other clients. Cursor, VS Code, Gemini, Perplexity, and custom agents connect to https://mcp.semrush.com/v2/mcp with an API key in the Authorization header. The developer docs list the config for each one. For terminal work, Claude Code with Semrush applies the same idea with more automation on top.

    Fourth, confirm the connection by asking something cheap, like the Semrush Rank for a domain in the US database. If a number comes back, the connection is live. If there is an error, ask the AI to walk through fixing it.

    Every prompt below sits in the Semrush MCP prompt library, where each use case is a workflow of three or four chained prompts. The first prompt in each workflow is featured here because it pulls the data; the full workflow link covers the follow-ups. To use one, paste it, swap in the relevant domain, country, and keywords inside the brackets, run it, and read the output before acting on it.

    The MCP pulls live Semrush data on demand and does not monitor anything or send alerts. Anything that needs watching over time, such as rank tracking or competitor alerts, gets set up inside Semrush itself. The MCP is also read-only, so it retrieves data but never changes it.

    What keyword strategy workflows does the Semrush MCP unlock?

    These search-demand and keyword-strategy prompts uncover where demand in a niche actually sits, which keywords competitors own that a site does not, and which gaps deserve effort first.

    Spot shifts in search demand

    This prompt maps the niche’s biggest keyword clusters by combined volume before the full workflow layers rising, declining, and SERP-opportunity views on top. Use it when planning a quarter and need to know where demand lives before deciding what to build.

    Using Semrush keyword data for {country}: Identify the top 8 keyword clusters for {niche} by combined monthly search volume. Return ONE table: Columns: cluster_name, combined_monthly_volume, example_keywords (up to 5). Limit: 8 clusters exactly.

    The output is a table of eight clusters ranked by combined volume. The workflow’s trend prompts then show which clusters are growing.

    Turn keyword gaps into roadmaps

    This prompt finds keywords competitors rank for that the site does not, plus the ones where the site ranks far behind. The full workflow then clusters them and plans pages. Reach for it when traffic is going to competitors but not through obvious doors.

    If {competitor-domains} are provided, use them directly (up to 5). If not, first find {your-domain.com}’s top organic search competitors (limit 5); exclude domains with Competitor Relevance = 0.00 and organic traffic above 10M (e.g., YouTube, Reddit, Wikipedia). Use Semrush data for {country} to analyze {your-domain.com} against its top organic competitors. Find and prioritize two opportunity types: Missing keywords (competitors rank, but {your-domain.com} does not) and Weak shared keywords (both rank, but {your-domain.com} ranks much lower than the strongest competitor). Prioritize low-hanging fruit that look actionable through content or on-page improvements. Return ONE table (up to 50 rows): Columns: keyword, opportunity_type, monthly_volume, intent, top_competitor_domain, competitor_rank, your_rank, rank_gap, recommended_action, rationale.

    The result is a 50-row table splitting gaps into missing and weak shared, with a recommended action per row. The workflow’s next prompts cluster the list into themes.

    Prioritize gaps by demand and intent

    This prompt estimates the search intent mix inside each gap cluster so gaps can be sequenced by business value rather than raw volume. The full workflow carries it through competitor difficulty checks into a six-week sprint plan. Run it as a follow-up once a gap list exists.

    Using Semrush keyword data for {country}: For the 8 gap clusters, estimate intent distribution. If gap clusters already exist in this conversation, use them. If no gap clusters are available, first identify keyword gaps for {your-domain.com} against up to 5 competitor domains, then cluster them into exactly 8 themes. Return ONE table: Columns: cluster_name, informational_share_pct (est), commercial_share_pct (est), transactional_share_pct (est), top_intent_keywords (up to 5). Limit: 8 rows. If intent labels are unavailable, infer from SERP/page types and label as estimate.

    The output is an eight-row table of estimated intent shares per cluster. Spot-check a SERP or two before trusting the roadmap, since the intent figures are model-generated estimates.

    How does the Semrush MCP support competitive intelligence?

    These competitive-intelligence prompts identify who a site actually competes with in search, how traffic splits, who is growing, and how to watch them without living in dashboards.

    Identify true search competitors

    This prompt ranks the domains sharing keywords by Semrush’s Competitor Relevance score rather than by assumed competitors. The full workflow maps overlap clusters and SERP feature wins next.

    Using Semrush data for {country}: Identify the top 10 organic competitors of {your-domain.com}, excluding high-traffic generic domains (Competitor Relevance = 0.00 or organic traffic above 10M, e.g., YouTube, Reddit, Wikipedia). Return ONE table: Columns: competitor_domain, estimated_organic_traffic, ranking_keywords, keyword_overlap_with_{your-domain.com}, overlap_pct (if available). Limit: top 10 competitors.

    The output is a 10-row table with traffic, keyword counts, and overlap per domain. Those names feed the next four prompts.

    Prioritize strengths, gaps, and attacks

    This prompt finds clusters where the site is strong and competitors are weak, so the focus is on defending strengths before chasing new ones. The full workflow ends in a defend-versus-attack map.

    Using Semrush data for {country}: Identify keyword clusters where {your-domain.com} has relatively strong visibility but the top 5 competitors have weaker presence. If a previous competitor focus or cluster analysis exists in this conversation, use it as a starting point. Return ONE table: Columns: unique_cluster, why_unique (1 sentence), example_keywords (up to 5), suggested_defense_action. Limit: 8 clusters.

    The output lists up to eight clusters where the site leads, each with a defense action. The workflow’s next prompts then show where competitors outrank the site.

    Size traffic share across competitors

    This prompt pulls each domain’s traffic and engagement and computes market share. It calls Traffic Analytics, which needs a Trends API subscription; the full workflow continues into channel and geography splits.

    For each competitor domain ({competitor-domains}) in {country}, use Semrush Traffic Analytics to retrieve each domain’s overall traffic summary (visits, unique visitors, engagement). It accepts multiple domains per request. If {competitor-domains} are not provided, use competitor domains identified earlier in this use case as top organic/search competitors, market competitors, or strongest keyword-overlap competitors. Combine, dedupe, and limit to 5. Request columns: target, rank, visits, users, pages_per_visit, bounce_rate, time_on_site. Build a single comparison table: Domain, Rank, Visits, Unique Visitors, Pages/Visit, Bounce Rate, Avg Duration (s), Traffic Share %. Derive: total_market_traffic (sum of all visits), market_leader (domain with highest visits), traffic_concentration (combined traffic share of top 2).

    The result is a share-of-market table with total market traffic, the market leader, and top-two concentration. On a plan without the Trends API, the MCP reports the gap and falls back to organic estimates, so a low_data engagement column usually means the plan, not a broken prompt.

    Find competitors gaining organic traffic

    This prompt pulls 12 months of traffic history per competitor and ranks them by absolute growth. The full workflow turns the winners’ patterns into playbooks.

    Using Semrush data for {country} over the last 12 months, identify the top 10 competitors of {your-domain.com} by organic traffic growth. Process: find the top organic competitors, select top 10 by keyword overlap or competitive relevance (exclude domains with Competitor Relevance = 0.00 and Organic Traffic above 10M). For each competitor, pull its organic traffic trend over time. Extract traffic_12m_ago and traffic_now. Compute traffic_change_abs and traffic_change_pct. Flag any competitor where traffic_12m_ago is below 100 as low_data, since percentage growth from a tiny base is misleading. Determine the top_growth_cluster driving traffic growth. Return ONE table: Columns: competitor_domain, traffic_change_abs, traffic_change_pct, top_growth_cluster. Limit: top 10 by traffic_change_abs.

    The output is a growth leaderboard with the cluster driving each gain. The low_data flag matters, because percentage growth from a tiny base will otherwise top the table.

    Track competitor visibility shifts

    To track visibility shifts, set the watch list first. This prompt builds a monitoring table of the 10 most relevant competitors and the cluster each competes on. The full workflow establishes the baseline that shifts are measured against.

    Using Semrush data for {country}: Identify the top 10 organic competitors of {your-domain.com} to monitor. Sort by competitor relevance (Cr) descending. Return ONE table: Columns: competitor_domain, estimated_organic_traffic, keyword_overlap (if available), primary_competing_cluster. Limit: 10 competitors.

    The result is a compact watch list. Regenerate the list monthly and hand it to the alerts prompt below.

    Build alerts and response plays

    This prompt produces alerts for when a competitor gains rankings or when the site loses them. The MCP writes the rules but cannot create alerts; the output goes into Semrush, for example as Position Tracking campaigns. The full workflow adds response playbooks.

    Create a competitor monitoring alert ruleset for {your-domain.com} in {country}. Do NOT generate alerts for {your-domain.com} gains; the purpose is early warning, not reporting success. Cover two signal categories: competitor_gain (a monitored competitor gains organic traffic, rankings, or visibility above threshold in clusters overlapping with {your-domain.com}) and own_loss ({your-domain.com} drops in rankings, traffic share, or keyword visibility in a monitored cluster). Return ONE table: Columns: alert_name, signal_type (competitor_gain / own_loss), metric, threshold, cadence, action_owner_role, what_to_investigate. Include at least 8 alert rules: minimum 5 of type competitor_gain, minimum 2 of type own_loss.

    The result is a rules table with thresholds, cadences, and owners. If the prompt runs in the same conversation as the earlier competitive prompts, the AI calibrates against real baselines in the conversation and every threshold carries both a percentage and an absolute floor instead of a generic number. Backtest it against a prior year’s data before accepting the output.

    FAQ

    What is the Semrush MCP server?

    The Semrush MCP server connects Semrush data directly to AI assistants such as Claude and ChatGPT, so users can run real SEO research in plain language through the Model Context Protocol (MCP), an open standard created by Anthropic that lets AI models connect to external data sources, files, and tools.

    Which Semrush plans include MCP access?

    MCP access comes with Semrush One Starter, Semrush One Pro+, SEO Classic Pro, and SEO Classic Guru, each including 50,000 API units. Traffic Market reports need a separate Trends API subscription.

    Which AI clients support the Semrush MCP endpoint?

    Claude and ChatGPT connect through their built-in connectors using OAuth. Cursor, VS Code, Gemini, Perplexity, and custom agents connect to https://mcp.semrush.com/v2/mcp with an API key in the Authorization header.

    Does the Semrush MCP monitor or send alerts?

    No. The MCP pulls live Semrush data on demand and does not monitor anything or send alerts. Anything that needs watching over time, such as rank tracking or competitor alerts, must be set up inside Semrush itself.

    Try the rank tracker

    SEOScanPro, which includes the rank tracker

    The rank tracker runs a full technical audit of a site and shows the measured result behind every check. Open the rank tracker.

    This article summarizes reporting from semrush.com.

  • Google Says EU DMA Changes Led to the Worst Search Quality Drop It Has Seen

    Google Says EU DMA Changes Led to the Worst Search Quality Drop It Has Seen

    Readers who rely on Google’s results inside the European Union are now navigating the steepest drop in search result quality that the company has recorded since the current measurement approach began. New data shared by Google ties that decline directly to changes made to comply with the European Union’s Digital Markets Act.

    What changed in EU search results

    Google has been running a series of experiments to meet its obligations under the Digital Markets Act, the EU regulation that forces large platforms to open up choice for users. Under those tests, Google has been surfacing results from rival vertical search engines and has been allowing those competitors to appear alongside, or in place of, its own specialized results such as flights, hotels, shopping and local listings.

    The trade-off, Google argues, is visible quality. The company says the EU versions of its search results now show the largest amount of low-quality, unredacted content it has ever observed in any of its markets.

    Why Google says the results are worse

    Vertical search engines and aggregators that are now being elevated inside Google’s EU results vary widely in how they source, verify and present information. Some pull data from user reviews, third-party feeds or loosely moderated databases. Others specialize in narrow commercial niches where accuracy matters more than breadth. When these sources are placed into the same result page as Google’s own systems, which blend ranking signals, spam detection and policy enforcement, the overall mix shifts.

    Google has framed the issue as a direct consequence of regulatory design rather than a choice by the company. The Digital Markets Act, in Google’s reading, leaves less room for the company to filter or rank down lower-quality sources, because doing so risks favoring its own properties over rivals.

    How the company measured the drop

    Google has used internal search quality evaluation processes, the kind its engineers run continuously across all markets, to compare the content surfacing inside the EU against the same queries in regions where DMA-style obligations do not apply. The output of those evaluations, according to Google, shows EU results producing more unredacted, lower-quality content than any other region it tracks.

    What this means for users and businesses

    For users searching in the EU, the practical effect is a result page that contains a higher share of content from sources that may not meet the bar Google’s own ranking systems would otherwise apply. For businesses that depend on Google’s organic traffic, the implication is that the competitive field inside EU results has widened, with more vertical players and aggregators able to occupy positions that Google’s own properties would previously have held.

    The wider question the data raises is whether the DMA’s competition goals and the quality goals that search engines are designed to serve can be balanced without one eroding the other. Google has made clear, through this report, where it believes the current balance sits.

    FAQ

    What is the Digital Markets Act?

    The Digital Markets Act is an EU regulation that requires large online platforms, known as gatekeepers, to follow rules that promote competition and user choice, including how they display results from rival services.

    How has Google changed EU search results under the DMA?

    Google has run experiments that include results from rival vertical search engines and has given those services placement alongside or instead of its own specialized results in categories such as flights, hotels, shopping and local.

    What is Google measuring when it talks about search quality?

    Google runs ongoing internal search quality evaluations across regions and compares the amount of low-quality, unredacted content surfacing in EU results against other markets, and reports that the EU shows the highest level it has ever recorded.

    This article summarizes reporting from searchengineland.com.

  • Google Page Experience Docs Add CrUX Ad Metrics

    Google Page Experience Docs Add CrUX Ad Metrics

    Site owners evaluating page experience now have new Chrome-reported metrics for understanding how ads affect real users. Google updated its page experience help document on September 22, 2026 to include experimental CrUX ad metrics, which measure ad count, ad density, and the resources ads consume.

    The new line appears in the ‘Page experience resources’ section of the document. It reads: ‘CrUX ad metrics: These experimental CrUX metrics from Chrome can help you understand your site’s ad count, ad density, and ad weight metrics. They provide insight into real users’ ad experiences on your site, as seen and reported by Chrome.’

    Google has not stated that these metrics are a direct ranking signal for search. The clarification from Google was direct: the metrics were included because they are available as tools to help evaluate the state of a site. Not everything is a direct search ranking factor, and having metrics can make it easier to discuss and work on site experience regardless of whether it falls under SEO.

    What Does Each New CrUX Ad Metric Measure?

    Four distinct measurements are now documented. Each captures a different aspect of how ads impact the real-world browsing experience.

    • Ad Count: The average number of ads in the viewport.
    • Ad Density: The average fraction of the viewport area occupied by ads.
    • Ad Weight (Network Usage): The resources consumed by ads measured in bytes.
    • Ad Weight (CPU Usage): The resources consumed by ads measured in milliseconds.

    These metrics come from CrUX, the Chrome User Experience Report, which collects data from real Chrome users who have opted in. Because they reflect actual browsing sessions rather than lab simulations, they show what people encounter on a live site.

    Why Page Experience Metrics Cause Confusion

    The addition arrives amid ongoing debate about how much page experience and Core Web Vitals influence search rankings. Google’s statements on the subject have shifted over time. At one point the company said it did not confirm page experience or Core Web Vitals as direct ranking factors. Later, Core Web Vitals were confirmed as a ranking factor.

    The confusion traces back to changes made about a year ago to the helpful content guidance and page experience documentation. Google later clarified that page experience is a ranking signal but not a ranking system. Signals are used by other systems, while systems are the mechanisms that actually determine rankings. Google has also removed the page experience report from Search Console, adding to the uncertainty for site owners trying to interpret what matters.

    The key takeaway for the new ad metrics is that they are diagnostic tools, not ranking levers. A site with poor ad experiences may still benefit from reviewing these numbers, because they provide a shared language for identifying and fixing problems.

    How Site Owners Can Use These Metrics

    The value of the CrUX ad metrics lies in measurement rather than direct ranking impact. A publisher can see how many ads appear in a typical viewport, how much of the screen they cover, and whether they drain network or CPU resources. That data can guide decisions about ad placement, lazy loading, or reducing heavy ad scripts.

    These measurements align with the broader principle that improving a website is not only about ranking factors. Metrics make problems visible and measurable, which enables concrete work on site quality. Where ad weight is high, performance work becomes a matter of reducing bytes and milliseconds rather than guessing.

    FAQ

    Are the CrUX ad metrics a Google ranking factor?

    No. Google did not state that the CrUX ad metrics are a direct ranking signal for search. They were added to the page experience documentation because they are available metrics to help evaluate the state of a site.

    What do the new CrUX ad metrics measure?

    They measure four things: the average number of ads in the viewport, the average fraction of the viewport occupied by ads, the resources consumed by ads measured in bytes, and the resources consumed by ads measured in milliseconds.

    Where can I find the CrUX ad metrics in Google’s documentation?

    The metrics appear in the ‘Page experience resources’ section of the page experience in Google Search results help document.

    Try the site audit tool

    The SEOScanPro site audit report

    The site audit tool runs a full technical audit of a site and shows the measured result behind every check. Open the site audit tool.

    This article summarizes reporting from seroundtable.com.

  • Claude Opus 5.5 Cuts Costs and Adds Safeguards for Autonomous AI

    Claude Opus 5.5 Cuts Costs and Adds Safeguards for Autonomous AI

    Developers gain a model that finishes multi-hour engineering work in roughly one-seventh the time of its predecessor while spending far fewer tokens. Claude Opus 5.5, released by Anthropic, is now available across the Claude Platform, Amazon Web Services, Google Cloud and Microsoft Azure under the model name claude-opus-5-5. The release pairs lower API pricing and faster output with watermarking designed to comply with the EU AI Act and new safeguards for sensitive work.

    Built for Long and Complex Tasks

    Opus 5.5 is designed for codebase migrations, software audits, financial analysis, data collection and workflows that span several applications. Early testers put it to work on engineering jobs that ran for hours at a time. In one documented test, Opus 5.5 audited and fixed a 200,000-line codebase in under three hours. Opus 5 needed more than 20 hours for the same task and used 2.5 times as many tokens.

    Clio, a legal technology company, assigned the model a large engineering task across six repositories and let it run overnight unattended. The model stayed on task for over 18 hours, defining how the services communicate and working out how each one should apply the result. The account from Clio’s development team was that milestones arrived faster than with Opus 5 and required minimal reworking. The code comments were short and useful rather than long and prose-heavy.

    Terminal-Bench 4.0, a benchmark that measures how well a model completes complex multi-step professional tasks inside a command line interface, was among the evaluation tools cited. Early testers also reported improvements in maintaining context, delegating work to other agents and checking results. Their accounts suggest the model can reduce the number of prompts, tool calls and corrections needed to finish a task.

    What Changed in Coding and Knowledge Work

    Anthropic reports gains in coding, computer use and professional knowledge work. Benchmarks and customer tests indicate that Opus 5.5 completes many tasks with fewer tokens than Opus 5, though performance varies according to the task, tools and effort settings.

    Deloitte Consulting tested the model on code review and US consulting analysis. At its lowest effort setting, Opus 5.5 caught 72% of known bugs in code reviews, while Opus 5 at high effort caught 56%, with fewer false alarms and a fraction of the output. On consulting analysis, low thinking effort matched the higher thinking settings on half the output and passed Deloitte’s quality checks, which means more lower-effort deployments can be put into production and still deliver client-ready work.

    Thinking Mode Stays On

    Opus 5.5 cannot be used with thinking switched off. The model always runs a reasoning process, and developers can adjust how much effort it applies to a task. Anthropic says the model is easier to understand, places key information earlier in responses and follows writing instructions more closely.

    The release also introduces preserved thinking, a safeguard that stops API users from editing the model’s prior context. Anthropic says the measure makes it harder to extract and copy the model’s capabilities through large-scale distillation attacks. Preserved thinking applies to Opus 5.5 API accounts created on or after August 31, 2026.

    Additional Controls for Sensitive Work

    Opus 5.5 includes safeguards covering cybersecurity, biology and attempts to copy the model. Most cybersecurity tasks are rerouted to Opus 4.8, and Anthropic plans to expand its Cyber Verification Program to give verified cybersecurity professionals broader access to Opus 5.5. Organizations whose biological research is impeded by the safeguards can apply to the Life Sciences Verification Program.

    Anthropic added a classifier that screens coding-agent actions before execution. The company also introduced an open-source sandbox that security teams can audit, plus code-review features designed to catch vulnerabilities before changes are merged. The model has stronger defences against prompt-injection attacks. In an evaluation of attempts to cross containment boundaries, Opus 5.5 tried to circumvent its assigned limits about 85% less often than Opus 5 or Claude Mythos 5.1. Anthropic said every attempt was low severity and self-reported.

    External organizations, including METR and Frontier Design, evaluated the model before release. Anthropic acknowledges that current evaluations cannot identify every potential failure before deployment.

    Lower API Prices and Faster Output

    Claude Opus 5.5 costs $4 per million input tokens and $20 per million output tokens, down from $5 and $25 for Opus 5. Cached input reads cost $0.20 per million tokens, down from $0.50. The lower cache price benefits coding agents and other systems that consult the same instructions, files or conversation history repeatedly.

    The model generates output more than 30% faster than Opus 5. A separate fast mode is available through Claude Code and the Claude Platform, offering up to 2.5 times the speed for $8 per million input tokens and $40 per million output tokens. Anthropic is increasing five-hour usage limits for Pro, Max, Team and seat-based Enterprise customers. Subscription users will receive a rate-limit reset that they can save and use later.

    FAQ

    How much does Claude Opus 5.5 cost?

    Claude Opus 5.5 costs $4 per million input tokens and $20 per million output tokens, down from $5 and $25 for Opus 5. Cached input reads cost $0.20 per million tokens, down from $0.50. A fast mode costs $8 per million input tokens and $40 per million output tokens.

    How much faster is Claude Opus 5.5 than Opus 5?

    Opus 5.5 generates output more than 30% faster than Opus 5. Its fast mode offers up to 2.5 times the speed. In one test, Opus 5.5 audited and fixed a 200,000-line codebase in under three hours, while Opus 5 took more than 20 hours and used 2.5 times as many tokens.

    Can developers turn off thinking mode in Claude Opus 5.5?

    No. Opus 5.5 cannot be used with thinking switched off. The model always uses a reasoning process, and developers can adjust how much effort it applies to a task.

    This article summarizes reporting from helpnetsecurity.com.

  • Federal Register Used Chinese Qwen AI Search Tool, Then Removed It

    Federal Register Used Chinese Qwen AI Search Tool, Then Removed It

    The National Archives pulled an open-weight Alibaba Qwen AI search tool from the Federal Register website on Wednesday after social media users spotted the contradiction: a US government site was running a Chinese model while the FBI had just named Alibaba as a threat to American AI leadership. The model had been offered as an option for searching public comments on proposed regulations and was available for at least a day before being taken down.

    How the federal government stack ended up with Qwen AI search

    An archived snapshot of the Federal Register source code confirmed that the Qwen model was removed on Wednesday. A widely shared screenshot posted to X on September 15 by a user with the handle “tleilax___” showed the search option displayed on the government site. It is not clear when the National Archives, which runs the Federal Register, first offered the tool to visitors, and the agency has not commented on the removal.

    The White House and the FBI also did not comment.

    The contradiction the FBI had just warned about

    Earlier in September, the FBI named Alibaba among six leading Chinese firms it accused of conducting “industrial-scale distillation,” a practice the agency says helps China cut costs and shorten development time in the race for global AI leadership. That announcement made the appearance of an Alibaba product on a government website a glaring mismatch.

    Daniel Castro, president of the Information Technology and Innovation Foundation, called it an “insane” disconnect for a US agency to use an Alibaba model while the FBI encourages stakeholders to use only American models.

    What the model on federal register was doing

    The model in use was a small open-weight release identified as the Qwen3 0.6B level, according to a Chinese news report. It is not Alibaba’s largest flagship model. The report described it as serving solely to retrieve documents, providing no complex reasoning, and sending no government data to Alibaba or to any third party.

    For US agencies, downloading and running this kind of model locally can give the government more control over data than sending queries to larger models hosted externally. Security experts cited that ability, along with lower and more predictable pricing, as reasons such models have become a “default” AI search tool.

    Georgetown University Law Professor Anupam Chander noted that the Federal Register’s content is “already public, so the model was not working with sensitive government information.” Senator Mark Warner (D-Va.) said the security picture depends on whether any US data was processed by “Alibaba-controlled systems.”

    Lawmaker response: no Chinese models in federal government

    US Rep. John Moolenaar (R-Mich.), who chairs the House China Committee, said “no federal government entity should use a Chinese AI model.” “Doing so only makes the federal government more dependent on Chinese AI models, and that is not in the national interest,” Moolenaar said.

    Why the open-weight gap matters for US AI policy

    Policy research submitted to the US-China Economic and Security Review Commission in March warned that the United States may already be at a disadvantage because it focused too heavily on maintaining a lead in frontier AI while failing to advance its own open-weight ecosystem. China, the researchers wrote, recognized that gap and moved to fill it.

    US export controls are calibrated to constrain frontier training by restricting access to advanced semiconductors. They do not address the small-model deployment cycle, which requires less advanced compute, draws on openly available base models, and generates advantage through application rather than pre-training.

    If the models that matter most for industrial AI are small, specialized, and open, the current US policy framework could be targeting the wrong layer of the competition. Companies like Nvidia and Meta have vowed to help the United States catch up. Researchers have warned that if the United States loses a broad user base to China, it could set back America’s ability to set the technical standards and norms that will govern AI development for years to come.

    What is open-weight AI?

    Open-weight models are AI models whose trained parameters are published openly, so any organization can download them and run the model on its own hardware. Running the model locally means user queries never leave the operator’s own systems, which is one reason federal, state, and business users have gravitated to small open-weight releases for routine search and retrieval tasks.

    FAQ

    What AI model did the Federal Register use?

    The Federal Register briefly used a small open-weight Alibaba Qwen model at the Qwen3 0.6B level to search public comments on proposed regulations. It is not Alibaba’s largest flagship model and does not provide complex reasoning.

    Why was the Qwen model removed from the Federal Register?

    It was removed on Wednesday after social media users noticed the contradiction of a US government site using a Chinese AI model while the FBI had just named Alibaba among firms accused of industrial-scale distillation.

    Did the Qwen model pose a national security risk on the Federal Register?

    Experts told Reuters it is unlikely the deployment posed any substantial risk because the site’s content is already public and the model was running locally rather than sending queries to Alibaba systems. Senator Mark Warner said the risk picture depends on whether any US data was processed by Alibaba-controlled systems.

    BizScoreAI

    BizScoreAI, which includes the AI visibility scan

    BizScoreAI has the AI visibility scan scores how visible a business is to AI search and shows what its listing looks like to the engines people ask. Open the AI visibility scan.

    This article summarizes reporting from arstechnica.com.

  • What 25 Years of SEO Teach About Getting Found in AI Search

    What 25 Years of SEO Teach About Getting Found in AI Search

    SEOScanPro is built by NewSunSEO, which has been doing SEO in one form or another since 2001. Twenty-five years is long enough to watch search change shape many times, and each change taught the same lesson from a new angle: the sites that keep winning are the ones search engines can read, trust and quote without effort. AI search is the newest version of that lesson. It rewards the same foundations, and it asks for more work on top of them.

    The foundations that have held since 2001

    Every era of search has come down to the same sequence. A crawler finds the page, renders it, understands it, and decides whether it answers the question. The tools and the ranking signals changed constantly. That sequence did not.

    A site that loads quickly, states clearly what each page is about, links its pages together sensibly and keeps its technical house in order has had an advantage in every version of search we have worked in. Those same qualities are what let an AI assistant read a page and lift an accurate answer from it.

    What AI search adds on top

    Traditional search ranks pages. AI assistants assemble answers, and they name only a handful of sources in each one. That changes what visibility means in three practical ways.

    • Access comes first. Each assistant crawls with its own named agent, and a robots.txt file can allow one while blocking another. A site can rank well in Google and still be invisible to an assistant it has turned away without realizing it.
    • Clarity earns the quote. Assistants quote pages that state things plainly, with a clear question and a direct answer underneath it. Structured data that says what a business is, where it operates and what it offers gives them facts they do not have to guess at.
    • Consistency across the web counts for more. An assistant cross-checks what your site says against what other sources say about you. Business details that agree everywhere make you easier to identify and easier to recommend.

    Why AI visibility is a whole-business effort

    This is the part that has changed the most. For most of search history, SEO could deliver results largely on its own: fix the site, publish better pages, earn links. AI visibility draws on work that sits with other people in the business.

    Developers control whether pages render cleanly and load quickly. Whoever manages listings controls whether your name, address and phone number agree across directories. Content owners control whether pages answer real questions in the words customers use. Customer service shapes the reviews an assistant reads. The businesses that do best in AI answers are the ones where all of these people work from the same picture of what needs fixing.

    That shared picture is the real value of an audit. It turns a broad goal like “show up in AI answers” into a specific list that each person can act on.

    Measure what AI search actually uses

    A single average rank hides most of what matters now. The measurements that show where you really stand are more specific:

    • Which AI crawlers your robots.txt lets in, and which rule decides each one.
    • Whether your structured data is complete, or missing properties that cause it to be ignored.
    • How quotable your key pages are, based on how they are built.
    • How visible you are across your whole service area, not just at one address.

    SEOScanPro runs more than 85 checks across 17 scored categories and shows the measured value behind each one, so every finding comes with the evidence you need to act on it. The free AI Crawlability Test, Schema Markup Checker and AI Overview Readiness check each answer one of those questions on its own.

    FAQ

    Does traditional SEO still matter for AI search?

    Yes. AI assistants rely on the same things search engines always have: pages they can reach, render and understand. Strong technical foundations remain the starting point, and AI visibility builds on top of them.

    What is the first thing to check for AI visibility?

    Check whether AI crawlers are allowed to read your site at all. Each assistant uses its own named agent, and a single robots.txt rule can block one without affecting the others.

    Why does AI visibility involve more than the SEO team?

    Assistants weigh page speed and rendering, consistent business details across the web, clear answers in your content and what reviews say about you. Those sit with developers, listing managers, content owners and customer service, so the best results come when everyone works from one list of fixes.

    Try the site audit tool

    The SEOScanPro site audit report

    The SEOScanPro site audit runs a full technical audit of a site and shows the measured result behind every check. Open the site audit tool.

  • MacBook Uses Webcam, Mirror, and AI Agent to Code Its Own AMD GPU Drivers

    MacBook Uses Webcam, Mirror, and AI Agent to Code Its Own AMD GPU Drivers

    Older Intel MacBooks can now refine their own AMD Radeon support at the OS level, because a coding agent reads the live screen through a mirror to judge each driver change in real time. That setup was shown off this week on Omarchy, an agent-focused Linux distribution, and it points to a workflow where the machine verifies its own progress on a graphical task without a human in the loop.

    What the setup actually looks like

    A photo circulated online showing a MacBook propped up with its webcam pointed at a mirror, which reflects the screen back into the camera. The laptop is running an Omarchy session that is, in turn, running a programming agent while it works on AMD Radeon driver tuning. Because the agent can see what its changes render on screen, it can confirm visually whether a tweak worked before moving on.

    Omarchy is described by its developers as a Linux distro “designed for the age of agents,” and it ships with built-in agents that help debug issues during installation and configuration. The page also highlights a fast installer and a “vibe your way through every alteration, tweak, or trouble” approach. The project is distributed under the MIT license.

    Why a mirror and a webcam?

    The MacBook needs the mirror trick because the screen cannot photograph itself directly. By aiming the built-in webcam at a small mirror angled toward the display, the camera captures whatever the agent has just output, including driver logs, error dialogs, and frame-rate or render artifacts in whatever application is open. The agent then reads those images and decides what to change next.

    This kind of visual self-feedback has value when the work is graphical: driver tuning, UI layout, rendering quirks, and GPU-accelerated effects are easier to verify by looking than by reading log lines alone. A purely text-based coding loop can write code, build it, and run tests, but it can struggle to notice that a window is clipped, a shader is producing the wrong colour, or a driver change caused a tear in the output. A camera pointed at the screen closes that gap.

    How Omarchy fits into the picture

    Omarchy is built to be friendly to older hardware. Its developers list support for older Intel-based Macs, Apple Silicon Macs, modern x86 PCs, and low-spec machines, including a 2011 ThinkPad X220 with 2GB of RAM as a reference “potato PC.” Specialised drivers and configurations for those older Intel Macs are part of why an agent might be tuning Radeon support in this exact environment: it is hardware that still benefits from driver work, but is rarely the focus of large upstream investments.

    Because Omarchy treats AI agents as first-class users rather than add-ons, an agent can drive the install, fix issues as they come up, and now, as this demo shows, watch its own work. The mirror-and-webcam trick is not a product feature of Omarchy itself; it is a workaround the developer improvised so the agent running on Omarchy could close the visual feedback loop on a laptop with no second monitor.

    What this signals for agent-driven workflows

    Self-monitoring has been a missing piece in agentic coding. An agent that only sees text can miss visual regressions in any task with a graphical surface, from driver tuning to web frontend work. Adding a camera, even one pointed at the laptop’s own screen with a mirror, gives the agent the same evidence a human developer would use.

    It also hints at where self-hosted AI work is heading. The photo shows a modest laptop acting as both the host for the work and the verifier of its own results, with no external server or remote desktop session in the loop. For hobbyists and small teams, that is an attractive shape: one machine, one agent, one cheap camera, an ordinary mirror, and a feedback loop that runs without supervision.

    What to watch next

    Open-source drivers for AMD Radeon GPUs on older Intel Macs are the kind of long-tail project that benefits most from this approach, because commercial investment is limited. If the technique holds up, expect to see the same mirror-and-webcam rig applied to other visual tasks: UI polish, compositor tweaks, wayland session debugging, and game launcher quirks. The combination of an agent-ready distro, a camera, and a screen the agent can see is enough scaffolding for a surprising amount of self-directed debugging.

    FAQ

    What is Omarchy Linux?

    Omarchy is a Linux distribution built “for the age of agents.” It offers a fast installer, built-in agents that help debug issues, and configuration aimed at older hardware, including Intel-based Macs, Apple Silicon Macs, modern x86 PCs, and low-spec machines. It is distributed under the MIT license.

    Why is a MacBook using a mirror and a webcam to code?

    A coding agent working on AMD Radeon driver tuning needed visual feedback on its own changes. The MacBook’s screen cannot photograph itself directly, so the developer aimed the built-in webcam at a small mirror angled toward the display. The camera then captured the screen, and the agent used those images to judge whether each driver tweak worked.

    Is Omarchy only for older Intel Macs?

    No. Omarchy supports Apple Silicon Macs and modern x86 PCs as well, in addition to older Intel-based Macs. The developers also highlight low-end machines such as a 2011 ThinkPad X220 with 2GB of RAM as supported hardware.

    This article summarizes reporting from tomshardware.com.

  • Google Brings Trip-Planning Features to Search AI Mode

    Google Brings Trip-Planning Features to Search AI Mode

    Travelers can now plan and book trips directly through Google’s AI Mode in Search, with new tools for tracking flight prices, booking hotels, and viewing costs in miles or points. The features went live on August 27, 2026, bringing the Google Flights price tracker, hotel booking with Google Pay, and rewards-point pricing into the chatbot experience.

    Flight price tracking comes to AI Mode

    The price-tracking tool from Google Flights is now built into AI Mode. You can tell the chatbot where you want to fly and when, and it presents options from more than 300 airlines and travel sites. If you then ask AI Mode to track prices, Google sends an email when prices change for your planned destination and travel dates.

    This feature is live in all locations and languages where AI Mode is available, except for countries and territories in the European Economic Area.

    Check flight and hotel costs in miles or points

    Users everywhere can now use AI Mode to check how much a flight or hotel costs in miles or points. Tell the chatbot where and when you want to travel and you will see the flight cost in miles. For hotels, tell AI Mode you want to see the cost in reward points.

    The feature initially works for options from a set of loyalty programs:

    • Alaska Airlines and Hawaiian Airlines
    • American Airlines
    • Choice Hotels International
    • Hilton
    • Wyndham Hotels & Resorts

    Google says support for Accor, Flying Blue, Hyatt, LATAM Airlines, and Lufthansa Group is coming soon.

    Book a hotel stay through AI Mode

    You can also book a hotel stay directly through AI Mode. After you enter details about your trip and preferences, Google displays hotel options alongside guest reviews. You can complete the booking using Google Pay.

    This feature initially launches in the US in English and works with partner sites including Expedia, Marriott International, and Priceline. Google plans to roll it out in the coming weeks.

    What this means for trip planning

    The changes fold several existing Google travel tools into one conversational interface. Instead of switching between Google Flights, hotel search, and loyalty program dashboards, you can ask AI Mode to find options, watch for price drops, and compare what a trip costs in cash versus points. The airline pool for miles pricing covers major US carriers at launch, while the hotel side starts with Hilton, Choice, and Wyndham before expanding to Accor, Hyatt, and Lufthansa Group properties.

    FAQ

    Can I track flight prices in Google AI Mode?

    Yes. Tell the chatbot where you want to fly and when, and it will show options from more than 300 airlines and travel sites. If you then tell AI Mode to track prices, Google sends an email when prices change for your destination and dates. This feature is live everywhere AI Mode is available except the European Economic Area.

    Can I book a hotel through Google AI Mode?

    Yes. After you enter your trip details and preferences, AI Mode displays hotel options alongside guest reviews. You can complete the booking using Google Pay. The feature initially works in the US in English with partners including Expedia, Marriott International, and Priceline, and rolls out more widely in the coming weeks.

    Which loyalty programs work with the miles and points pricing feature?

    At launch, the feature works with Alaska Airlines, Hawaiian Airlines, American Airlines, Choice Hotels International, Hilton, and Wyndham Hotels & Resorts. Support for Accor, Flying Blue, Hyatt, LATAM Airlines, and Lufthansa Group is coming soon.

    This article summarizes reporting from engadget.com.

  • Google Search Console Releases Generative AI Performance Reports and AI Controls to Everyone

    Google Search Console Releases Generative AI Performance Reports and AI Controls to Everyone

    Site owners now have a direct way to measure how often their pages appear in Google’s generative AI features and to control whether their content can be used for AI grounding. As of August 31, 2026, Google has rolled out its Generative AI performance reports and search AI controls in Google Search Console to all websites worldwide.

    The rollout began in the first week of June and expanded access gradually over the following months. Google confirmed the global availability in a blog post and help document, noting that if a site owner does not see the report, it may be because there is not enough data for that site yet.

    What the Generative AI Performance Report Shows

    The new report appears as an expandable tab under the main performance report in Google Search Console. It is designed to show how a site’s URLs perform specifically within generative AI features in Search and Discover.

    The report includes several data dimensions, though click data and query data are not included:

    • Impressions: How often URLs from a site appeared in generative AI features in Search and Discover.
    • Pages: Which URLs appeared within AI features.
    • Countries: Visibility broken down on a country basis.
    • Devices: The devices people are using when seeing the website, available for Search results.
    • Dates: Performance over time with hourly, daily, weekly, and monthly granularity.

    This gives site owners a clear view of whether their content is being surfaced in AI Overviews, AI Mode, and similar generative experiences, and where that visibility is coming from.

    How the AI Blocking Control Works

    Alongside the report, Google is rolling out a toggle that lets site owners block their content from appearing in or around AI features such as AI Overviews, AI Mode, and AI Overviews in Discover.

    The toggle controls whether a site’s content can be used in these AI features, either as links or for grounding. This gives publishers a choice about participation in generative AI search experiences without affecting their standard search listings.

    What This Means for Site Owners

    For site owners tracking visibility in AI-driven search results, this report closes a measurement gap. Until now, there was no direct way in Search Console to see impressions generated specifically within AI features. The hourly to monthly date granularity also supports monitoring how AI visibility shifts after content changes, algorithm updates, or new AI feature launches.

    The device and country dimensions make it possible to spot where AI visibility is concentrated, which can guide content decisions for specific markets. For those who prefer not to have their content used in AI features at all, the new toggle provides a direct opt-out mechanism.

    FAQ

    What is the Google Search Console Generative AI performance report?

    The Generative AI performance report is a new expandable tab under the main performance report in Google Search Console. It shows how often URLs from a site appeared in generative AI features in Search and Discover, including impressions, pages, countries, devices, and dates.

    Does the Generative AI performance report include click data?

    No. The report does not include click data or query data. It focuses on impressions and visibility metrics such as pages, countries, devices, and date ranges.

    How do I block my site from Google’s AI features?

    Google has rolled out a toggle in Search Console that lets site owners opt out of their content appearing in or around AI features such as AI Overviews, AI Mode, and AI Overviews in Discover. The toggle controls whether content can be used for links or for AI grounding.

    Related coverage

    Try the AI visibility report

    SEOScanPro, which includes the AI visibility report

    The AI visibility report runs a full technical audit of a site and shows the measured result behind every check. Open the AI visibility report.

    This article summarizes reporting from seroundtable.com.

  • 20 SEO Questions, 20 Sharp Answers: A Practitioner Q&A

    20 SEO Questions, 20 Sharp Answers: A Practitioner Q&A

    Twenty focused questions produce twenty concrete answers about where SEO is heading, what changed when AI search engines entered the picture, and how practitioners are reorganizing their work in response. The exchange covers site migrations, link building, brand mentions, analytics, and the practical overlap between traditional search optimization and getting cited by AI assistants.

    What did the interview actually cover?

    The format was deliberately tight: twenty questions, twenty answers, no filler. The interviewer pushed for specifics on technique, measurement, and prioritization, and the answers stayed grounded in what practitioners see in day-to-day work rather than abstract trend pieces.

    Several threads ran through the conversation. One was how classic SEO fundamentals (crawl efficiency, internal links, clean redirects, accurate analytics) have become more important now that AI systems are pulling from the open web to build their answers. Another was how teams are deciding where to spend their limited bandwidth, especially when a single visibility report no longer captures the full picture.

    How is AI search changing SEO priorities?

    The clearest theme across the answers was a shift from a single ranking position to a broader concept of visibility. When a user asks an AI assistant a question, the assistant picks from sources it has indexed and synthesizes an answer. Showing up in that synthesized response is a different goal from ranking first in a list of blue links, and the answers described practitioners layering that goal on top of their existing work rather than replacing it.

    Brand mentions, even without hyperlinks, came up as a signal worth tracking. So did the structure of content: clear answers to clear questions, factual claims that can be lifted cleanly into a generated response, and on-site markup that helps machines identify what a page is actually about.

    What about site migrations and analytics?

    Migrations came up as one of the highest-risk moments for organic traffic, and the answers emphasized pre-migration baselines, mapping every redirect, and watching log files rather than trusting dashboards alone in the weeks after launch. Analytics hygiene showed up repeatedly: clean filters, correct hostname settings, and a working grasp of which queries and pages actually drove results before any change.

    How are teams approaching link building now?

    The answers treated link building as still relevant but repositioned around relationships, original data, and genuinely useful tools or research that other sites want to reference. Cold outreach for the sake of a link was described as a poor use of time. Digital PR, original surveys, and free tools were named as the kinds of assets that earn coverage organically.

    What practical advice came out of it?

    Several answers converged on the same practical point: measurement has to catch up with where users actually see a brand. A traditional rank tracker still has value, but visibility now also includes whether a brand is cited by AI systems, whether its pages are being pulled into AI overviews, and whether its name appears in the sources an assistant draws from. Tools that report a position on a service-area grid map the real-world picture more accurately than a single city-wide average, and SEOScanPro’s GEO Grids do exactly that kind of local rank mapping for businesses that need to see which towns and suburbs they are invisible in.

    Why does the format matter?

    Twenty short questions force short, specific answers. That format cuts through the broad trend pieces and leaves readers with a checklist of concrete shifts they can apply to their own work, from how they structure a piece of content to how they measure whether it landed.

    FAQ

    What changed in SEO because of AI search?

    Visibility now extends beyond a single ranking position. Practitioners track whether a brand is cited in AI-generated answers, whether its pages are pulled into AI overviews, and whether its name appears in the sources an assistant draws from, layering this on top of traditional ranking work.

    Are link building and brand mentions still important?

    Yes. Links from credible sources still matter, and brand mentions without hyperlinks are treated as a separate signal worth tracking. The strongest results come from original research, useful tools, and digital PR rather than cold outreach.

    What is the highest-risk moment for organic traffic?

    Site migrations. The answers stressed pre-migration baselines, full redirect mapping, and watching log files in the weeks after launch rather than relying on dashboards alone.

    Try the AI visibility report

    SEOScanPro, which includes the AI visibility report

    The AI visibility report runs a full technical audit of a site and shows the measured result behind every check. Open the AI visibility report.

    This article summarizes reporting from searchengineland.com.

  • Google Merchant Center AI Performance Report Gains Search Intent, Terms, and Attributes

    Google Merchant Center AI Performance Report Gains Search Intent, Terms, and Attributes

    Google Merchant Center AI Performance Insights now break down AI Search intent, AI Search terms, and AI attributes, giving merchants three new lenses on how their listings appear inside Google’s AI-powered results. The additions, spotted on the live report interface, build on a dashboard that first launched in July and has been expanding to more regions through early September.

    What is new in the AI Performance Insights report

    The AI Performance Insights report inside Google Merchant Center now includes three new reporting sections that look at how product data performs in AI-driven searches. The update was spotted in the live interface and screenshotted on X, where the observer confirmed the dashboard now groups its metrics under AI Search intent, AI Search terms, and AI attributes.

    Google revised its supporting help page for the report at the same time, publishing numerous changes that walk through the new sections and what each one measures.

    What each new section measures

    AI Search intent

    The AI Search intent section shows how products match customer AI searches. In practice, this is a view of whether the products a merchant is submitting line up with the queries that Google’s AI surfaces are running for shoppers.

    AI Search terms

    The AI Search terms section recommends improving visibility by including popular terms in product descriptions for items that already appear in AI searches. In other words, the report looks at which products are being picked up by AI results, then points to the wording that could be added to descriptions to push that visibility further.

    AI Attributes

    The AI Attributes section recommends increasing visibility by adding missing attributes to products that show up in AI searches. Where the terms section is about language, this section is about structured data: the colour, size, material, or category fields a product feed is missing, and that AI surfaces look at when deciding which products to return.

    What the changes mean for merchants

    Read together, the three additions shift the report from a single performance summary into a diagnostic tool. Merchants can see the gap between what AI search surfaces are looking for and what their product feed actually supplies, then close it with better copy or better attributes.

    For product feeds that already appear in AI results, the terms and attributes views point to the next step: more coverage and more relevance, rather than chasing entry into AI results in the first place.

    Where the report is available

    The AI Performance Insights report launched in July, and Google has been rolling it out to more regions since the start of September. The new intent, terms, and attributes sections are part of that same surface, so merchants who only recently gained access to the report should expect to see the additional sections appear as the wider rollout progresses.

    FAQ

    What is Google Merchant Center AI Performance Insights?

    AI Performance Insights is a Google Merchant Center report that looks at how products perform in Google’s AI-powered search results. It launched in July 2026 and has been expanding to more regions since early September, with new sections for AI Search intent, AI Search terms, and AI attributes added in September.

    What do the new AI Search intent, terms, and attributes sections show?

    AI Search intent shows how products match customer AI searches. AI Search terms recommends popular terms to add to product descriptions for items already appearing in AI searches. AI Attributes recommends missing attributes to add to those same products to increase their visibility.

    Why does the AI Performance Insights report matter for merchants?

    The report turns AI search visibility into something measurable and actionable. Instead of guessing why a product does or does not appear in AI results, merchants can see the intent behind the queries, the terms AI surfaces respond to, and the attributes still missing from the feed, and act on each one.

    Try the site audit tool

    The SEOScanPro site audit report

    The site audit tool runs a full technical audit of a site and shows the measured result behind every check. Open the site audit tool.

    This article summarizes reporting from seroundtable.com.