August brought a sharper view of where search is heading. AI visibility became easier to measure as Cloudflare and Microsoft launched new reporting tools, while OpenAI pushed ChatGPT Ads further into Europe. Social discovery also became more searchable, with Google extending Search Console reporting to off-site social content and TikTok giving creators more control over keyword metadata. At the same time, Reddit’s sudden 86% drop in ChatGPT citations was a reminder that visibility across AI platforms can change almost overnight. For brands, the challenge now is not simply expanding presence across more surfaces, but understanding which ones are driving real influence and how stable that visibility really is.
Here’s what changed across Google, AI, Social, and Paid and what it all means for brands navigating the next phase of search.
Google Search
Search Console Platform Properties Go Global
Google Search now supports platform properties in Search Console worldwide, helping creators and publishers track Instagram, TikTok, X, and YouTube posts across Search, Discover, and Google News. Google also published a new guide on how to analyze social and video content performance.
Why does it matter? In an era when publishers reach audiences through diverse channels beyond their own websites, these insights help teams understand how off-site social and video content performs in Google’s surfaces expanding visibility measurement well beyond traditional web pages.
Google Launches New AI Study Tools Across Search
Google announced a slate of new study tools across Search and Gemini on August 19, including AI-generated interactive visuals, 3D simulations, a dedicated student hub, and customized practice quizzes. On Search, students can now generate custom tools and simulations to understand complex topics for example, searching “pH scale” returns an interactive visual within an AI Overview.
Why does it matter? The launch marks Google’s latest effort to make Gemini the AI assistant students turn to, as it competes with OpenAI and education startups like Knowt and Gauth. It signals continued expansion of interactive, generative features directly inside search results.
Google Rolls Out the August 2026 Spam Update
Google released its third spam update of 2026 on August 18, with the rollout completing on August 21 after running roughly 2 days and 16 hours. It applies globally and to all languages, following the March and June rollouts. Google didn’t announce any new spam policies with this rollout, meaning existing spam policies remain the standard sites are measured against.
Why does it matter? Sites that violate Google’s spam policies may drop in rankings or fail to rank at all, with changes affecting Google Search, AI Overviews, and AI Mode results. Now that the rollout is closed, Search Console data from after August 21 gives a cleaner read for assessing impact.
Google Recommends Using 304 Status Code To Conserve Crawl Budget
Google updated its crawl budget documentation with new information about how Google’s different crawlers all share a website’s crawl capacity and added a recommendation to use a 304 server response to reduce the amount of server resources used by crawlers. Both of these additions are to Google’s Optimize Your Crawl Budget Documentation that is for enterprise websites with over a million pages and medium sized websites of over 10,000 web pages that rapidly change.
Why does it matter? If a site gets crawled by Googlebot-Image and Googlebot, both of the crawlers are sharing the crawl budget for the website. The impact is that a high rate of crawling by one bot can reduce the capacity available to other Google crawlers. The interesting recommendation that Google added was the suggestion to use 304 Not Modified server response code. A 304 Not Modified response tells Google’s crawlers that a page has not changed since the last time it was crawled. The result is that the server doesn’t serve the page to Googlebot and Googlebot can go focus on indexing other pages.
AI Search
Anthropic to watermark text generated by its AI models including Claude
Anthropic will watermark text generated by its models, including Claude, to comply with European regulations, the company now says. The AI model maker confirmed the watermarking in an updated support page. EU AI Act’s Transparency Code, which took effect on August 2, requires AI companies to mark AI-generated or edited content in a way other systems can identify them.
Why does it matter? If Google wasn’t able to detect AI content, it certainly will now – as will users. Although Google has not stated that AI content is worthy of penalisation, using Claude to create materials in violation of Google’s policies are. This watermarking will just make detection easier. Furthermore, brands who are using AI content must be prepared to own that decision. If brands are comfortable with users knowing they use AI, there will be no changes. Brands which aren’t may need to rethink their AI and Claude strategy.
OpenAI Expands ChatGPT Ads Into Europe
OpenAI continued its monetization push in August, expanding its advertising program. Brazil and Mexico went live on August 11, completing an initial rollout across nine markets, and the UK gained access to the self-serve Ads Manager beta that US businesses have had since May. The company formally announced “ChatGPT Ads expands across Europe” on August 19. Within roughly six weeks of the initial pilot, the ads effort was reportedly generating $100 million annualized, per an OpenAI spokesperson cited by Reuters.
Why does it matter? Ads inside AI answers change the economics of AI search and raise fresh questions for brands about visibility and placement. The Ads Manager also gained a Feeds section supporting up to 1 million SKUs per advertiser, signaling ChatGPT is building serious commercial search infrastructure to rival Google’s ad business.
AI Search Tools Still Struggle with Source Accuracy
Fresh August coverage renewed scrutiny of how reliably AI search cites its sources. A Columbia Journalism Review Tow Center study on AI-powered news search found these tools answered more than 60% of source attribution queries incorrectly across the board, with error rates ranging from 37% to 94% depending on the platform. Meanwhile, roughly 42% of US adults now use a chatbot specifically to search for information rather than a traditional search box, per Pew Research Center.
Why does it matter? As more users treat AI answers as their primary search interface, unreliable attribution has real consequences for publishers, brands, and anyone depending on AI for research. It’s the difference between an assistant you can lean on and one that confidently makes things up.
Meta/Facebook Is Crawling the Web and May Be Building Its Own Search Engine
Meta, the company behind Facebook, Instagram, WhatsApp, Messenger, and Threads, is allegedly crawling the web to build its own search engine. The claim came from Pieter Levels, who posted on X on August 6, 2026 after being tipped off by Meta staff and shared screenshots of Meta bots crawling his sites. The theory is that Meta wants its own web index so its AI can run web searches without routing through Google, which could otherwise use that activity for its own training. Facebook has chased search ambitions for over a decade, once powering its web search through a Bing partnership before dropping it a few years later. The move fits today’s AI landscape, where Google, Microsoft, and X are all building their own AI and can no longer rely on partners for a web index.
Why does it matter? If Meta builds its own index and search engine, it becomes another crawler webmasters need to account for and a possible new source of search and AI referral traffic outside Google. For an industry so dependent on Google, a serious Meta search play could change how sites approach crawl access, indexing, and AI search visibility. This remains unconfirmed and alleged for now.
Microsoft Clarity adds a new scrape-to-referral ratio metric
Microsoft Clarity now helps you connect AI crawling activity to actual customer sessions by combining Bot Analytics with Clarity’s behavioral analytics. With the new AI Scrape-to-Referral Ratio card and direct links into session recordings, you can evaluate the tradeoff between content extraction and traffic return, then investigate how AI-referred visitors behave once they arrive.
Why does it matter? The updated Microsoft Bot Analytics dashboard adds an AI Scrape-to-Referral Ratio and operator-level breakdown, letting users compare AI scraping activity against actual referral traffic to see which AI platforms return value versus which mostly extract content. Accuracy safeguards ensure ratios are calculated only on correctly mapped domains, with clear notices when bot and referral data coverage don’t align. Users can also jump directly from referral insights into filtered session recordings to validate attribution and assess true traffic quality—such as engagement, conversions, and bounce behavior—from each AI source.
Cloudflare Launches AEO Dashboard to Track AI Visibility
Cloudflare launched an AEO Visibility Dashboard, a new tool in its Answer Engine Optimization Suite that shows brands whether AI assistants are citing, mentioning, or recommending them in response to user queries. Unlike tools that rely on sampling test prompts sent to AI chatbots, Cloudflare draws on real crawl and referral data observed at the network layer across millions of sites it operates. The dashboard tracks metrics like Citation Rate, Mention Rate, Prominence, and Share of Voice, and pairs with Cloudflare’s existing Agent Readiness tool, which checks whether AI agents can even access and read a site in the first place.
Why does it matter? As AI assistants increasingly replace traditional search as a discovery channel, brands have had no equivalent of search rankings to know if — or how — they’re being surfaced in AI-generated answers. This tool gives marketers concrete, network-level data to diagnose specific problems (e.g., an “authority problem” vs. an “awareness problem”) and direct their content strategy accordingly. It also signals a broader shift in SEO-style tooling toward optimizing for AI agents rather than just search engines, which could reshape how businesses invest in content and visibility going forward.
Social Search
LinkedIn now allows users to flag posts as ‘AI slop’
LinkedIn is taking aim at low-quality, artificially generated content filling its feed. The company announced it’s adding a new feature to let users click a “seems like AI slop” button when someone’s post appears to have been written with AI. The move reflects a broader shift across online publishing platforms to cut back on AI content, as people have grown frustrated with the computer-written, inauthentic posts filling the web.
Why does it matter? If a LinkedIn account is flagged as ‘AI slop’, the ramifications are unknown but will likely result in a loss of platform visibility. The policing of this function (whether competitors can maliciously flag human-generated content as AI) is also unknown. Regardless, this does show that users desire human-led content on their social feeds.
Reddit Nearly Vanishes From ChatGPT Citations in an 86% Collapse
Reddit‘s share of ChatGPT Search citations fell 86.4% in four days, according to new Promptwatch data from an average of 3.83% between July 18 and August 7 down to 0.52% through August 17. The timing lines up with an August 8 change in ChatGPT’s query fan-out behavior, when its use of the “site:” search operator jumped from about 0.4% to nearly 17% of queries. OpenAI, however, says it does not target or set fixed visibility for individual websites and that Reddit continues to be cited.
Why does it matter? Communicators trying to influence what AI says about their brands are learning that generative engine optimization (GEO) is a rapidly moving target. A near-identical collapse happened in September 2025 and Reddit later recovered, so the drop is a signal to watch rather than a confirmed permanent shift but it shows how a single unannounced backend change can reshuffle social visibility overnight.
TikTok Lets Creators Manage Their Own Search Keywords
TikTok has made its shift toward search official by letting creators and brands manage the keyword metadata attached to their videos meaning they can remove keywords that don’t match their content and suggest ones that do, instead of leaving it entirely to TikTok’s automated systems. This signals something bigger: social platforms are actively building the infrastructure of search engines and asking creators to help train it with better data.
Why does it matter? Nearly half of US consumers have used TikTok as a search engine, per Adobe Express research, rising to 65% among Gen Z. These apps have evolved into full-blown discovery engines with their own ranking logic, so giving creators direct control over search metadata is a meaningful lever for brand discoverability.
Fresh Data Confirms Social Platforms as Primary Search Engines
Adobe‘s latest research quantified how far social search has come. Nearly half of consumers surveyed (49%) used TikTok as a search engine in 2026, up from 41% in 2024 a 19.5% increase in adoption in two years. Gen Z were also the most likely to find Reddit useful as a search engine (38%). Notably, though, among Gen Z the preference for TikTok over Google actually declined, from 8% in 2024 to just 4% by 2026.
Why does it matter? The data confirms social search is now mainstream for discovery, but nuances the “Gen Z is replacing Google” narrative: social platforms are a supplement to Google for certain visual, local, and recommendation-based questions, not a wholesale replacement. Brands need a presence on both surfaces.
Paid Search
Google Sets Migration Timeline for AI Max and Sunsets Legacy Broad Match
Google officially confirmed the roadmap to auto-migrate legacy Search campaign features to AI Max by September. Dynamic Search Ads (DSAs) and legacy Automatically Created Assets (ACAs) are moving toward intent-based targeting powered by landing page expansion and generative copy customization. Additionally, manual language targeting settings across Search and Performance Max (PMax) are being phased out in favor of automated auction-time language matching.
Why does it matter? This marks a major structural shift from keyword-led, manual targeting toward fully automated, intent-based decisioning. Media buyers must audit ad group landing page structures and language alignment ahead of September. Relying on tight broad-match constraints or legacy DSAs will no longer be an option; account performance will heavily depend on creative inputs, high-intent landing page signals, and first-party conversion feedback.
Microsoft Advertising Rolls Out AI Max Globally
Microsoft Advertising has begun a global rollout of AI Max for Search campaigns, moving it into general availability after an open pilot that started in May and the original April 21 announcement. AI Max adds three automation features to existing Search campaigns: search term matching (which reaches queries beyond an advertiser’s keyword list using signals from keywords, ads, landing pages, and user intent, including on Bing and Copilot), text customization (which generates extra ad messaging from existing assets and website content), and final URL expansion (which can swap in a landing page it judges a better match for intent). Advertiser controls are available from day one, including brand inclusions and exclusions, term exclusions for text generation, and URL rules to limit where final URL expansion sends traffic. Microsoft is folding Predictive matching and autogenerated text assets into AI Max, and campaigns already using either will have the matching setting turned on automatically, while other features stay off unless the advertiser opts in. AI Max settings can also carry over from imported Google Ads campaigns, with one exception: imported campaigns that originated as Dynamic Search Ads will be converted back to DSA, since Microsoft has not set a sunset date for DSA while Google already has.
Why does it matter? This gives advertisers a fuller keyword-light automation option on Microsoft Advertising and puts it more in step with Google’s push toward broader query matching. Because some AI Max settings may already be enabled in existing or imported campaigns, advertisers should review their accounts to confirm which features are active and set brand and URL controls before AI Max reaches into more of their traffic. The open question is whether the added automation delivers enough incremental value for advertisers to keep it on.
Meta’s New Delivery Engine Delivers Performance Gains—at the Cost of Creative Control
Meta has deployed its new Meta Generative Recommender model to optimize ad delivery, yielding an estimated 8.3% increase in ad clicks and a 15.7% conversion uplift on Facebook. Parallelly, Advantage+ now automatically modifies and rewrites headline text baked into creative assets unless brands explicitly set brand guardrails or opt out per creative.
Why does it matter? Meta’s algorithm is increasingly analyzing creative content end-to-end (visuals, overlay text, and landing page context) rather than relying solely on audience targeting signals. Advertisers running repetition-heavy variations without creative distinction will face higher CPMs and diminishing returns. Media teams must ensure visual identity rules are populated in Ads Manager to prevent unauthorised copy rewrites on key brand assets.
What Happened Last Month
In June, AI search moved beyond retrieval into action. Google doubled down on zero-click search, AI assistants became capable of shopping, booking and researching on users’ behalf, and platforms introduced the first real tools for measuring AI visibility. At the same time, social platforms strengthened their role as AI search sources, while paid media expanded further into conversational experiences with ChatGPT ads rolling out internationally and Google continuing to reshape advertising for AI-powered search. Across every channel, the focus is shifting from generating traffic to understanding influence, as AI continues to reshape how visibility is measured and valued.
Read the full July Search Update here.
