Agent Beat for October 15, 2026

Welcome to Agent Beat, the daily briefing on how AI finds brands and how agents choose what to buy. It is Thursday, October 15, 2026. Here is what happened.

OpenAI appears to have removed or replaced the UTM parameters it attached to links in ChatGPT results, according to Search Engine Roundtable. That makes it harder to tell whether a visit came from ChatGPT or another platform. The outlet says the change started with the GPT-6 release about a week earlier.

For site owners who want agents to do more than click, the Journal du Net explains WebMCP. It is a protocol that lets a site expose its own actions directly to AI agents instead of making an agent read and click through pages. The piece includes a tutorial. Marketers should check which actions on their site, such as search, booking or checkout, can be declared this way, because that is the surface agents will call.

The same readiness question applies to stores. Search Engine Journal tested 122 Shopify stores that were labeled agent-ready to see what that actually means. It found that a store can look agent-ready by default while failing as soon as a buyer asks about shipping, pickup or cancellation. The piece lists what to verify in store data before agents can act on it.

Visibility scores can hide a problem. MarTech reports that Semrush's 2026 AI Visibility Index looked at 22 categories across four AI platforms. It found that, on average, fewer than half of the top 100 mentioned brands also appeared in the matching top 100 citation lists in every month measured. The article separates a mention from a citation. A mention is when the brand appears in the answer. A citation is when its domain appears as a source. Ford is an example: it sat inside all four top-100 mention lists in the US in July, while its domain ranked as low as 1,320th for citations on AI Mode.

For procurement teams buying AI search work, iPullRank published a guide. It covers what AI Search is, the five names vendors use for it, GEO, AEO, LLMO, AIO and AI SEO, and why proposals are hard to compare. The post says AI systems judge a company on what the wider internet says about it, including press, reviews, Reddit and YouTube, as well as its own site.

On the advertising side, the agency Webloom tested ChatGPT Ads for eight advertisers. The Journal du Net reports that the format builds awareness but not profitability. The tests also produced biased measurement, according to the agency. Advertisers should treat these results as a reason to keep ChatGPT Ads at test budgets and to question any return figures a vendor shows them from early campaigns.

Buyers are changing how they buy software. According to G2's 2026 Buyer Behavior Report, more than 60% of B2B buyers use or plan to use AI agents in the software buying process, for research, shortlisting and comparison. People keep the final decision. MarTech writes that whether sales reps keep their jobs will depend on how far buyers trust their own agents to go. The headline says AI won't kill sales jobs, buyers will.

Search results themselves are also being studied. SISTRIX analyzed more than 10 million search results in seven countries. It found that the provider appearing most often, and most often at the top, in Google's EU comparison box is Google itself. Google introduced the box in the European Economic Area on September 8, replacing product carousels, after the European Commission fined it 460 million euros on July 23 over favoring its own services.

Marketing software is opening up to agents too. Positive says its marketing software now exposes MCP servers so AI agents can retrieve data, chain actions and run several marketing tools inside one conversation, according to Siècle Digital. The claim comes from the vendor.

Media companies are also building businesses around AI search. In the US and Europe, they are turning GEO into a product they sell to advertisers, according to Digiday Japan. Future says its GEO offering has more than 30 clients and renewals. Ziff Davis says it has several. One media executive said its GEO menu already brings in real revenue, though the figures are too early to size the market.

Advertising inside AI answers may grow. Google may show ads for certain pharmaceutical products inside AI Mode and AI Overviews. A Google spokesperson told Adweek that the company has recently expanded to include those products, Search Engine Roundtable reports. The report covers the ad surface, not the answers themselves.

Media organizations are also pushing back. More than 80 media organizations are uniting against the AI giants. Siècle Digital reports that The New York Times, The Economist and AFP have joined SPUR, a group of about 80 news organizations in more than 20 countries. The group wants to set the terms for AI companies using journalism and to make licensing deals easier. It added more than 40 members.

That is the briefing for Thursday, October 15, 2026. Thanks for listening to Agent Beat. For the full written brief, go to agent-beat.com.