Google's own monopoly judge just called AI Overviews 'really unfair' to publishers. That's the headline, but the real story is how every layer of marketing, search, content, creator, is being restructured around machines that cite strangers and demand new formats. The old playbooks are already losing.
Google hasn't broadly nuked listicles, but it has squeezed them out of the strongest spots. Across 60,000 US queries, listicles still made the top 10 in 55.1% of results, yet their share of the top three fell from 35.9% to 23.5%, per this study. The sharpest signal: self-serving brand lists are the casualty, with several SaaS sites losing 29-49% of organic visibility in weeks.
Amit Mehta, the same judge who found Google's search monopoly illegal, openly pressed back as Google argued at a Penske hearing that AI Overviews is just a "product improvement." Google's lawyers called the crawl-for-traffic arrangement with publishers a "vague historical course of dealing," while Mehta noted rivals pay for content and Google ingests the web free. He flagged that product improvements are not immune to antitrust scrutiny, a telling signal for every publisher eating lost clicks.
USA Today Co. is restructuring its content into machine-readable formats, markdown pages, metadata, modular templates, so AI crawlers can access, understand and cite it. The goal isn't just visibility in AI answer engines but a sales pitch: more licensing deals, on recurring terms. It's a clear bet that "good for the machine" now means good for revenue.
Across 1,851 citations drawn from ChatGPT, Perplexity, and Google AI Mode, only 2.8% pointed to brand-owned pages. Even when a tool recommended a brand by name, its own site was cited just 31% of the time. For ecommerce teams, being recommended and being cited are now separate outcomes worth tracking, and the gap is where visibility quietly leaks to third-party review sites.
A leaked Trade Desk roadmap reveals Ask Koa, a conversational interface that routes buyers to specialized AI agents handling audience building, campaign creation, insights and troubleshooting. The system ships with the Kokai Zuma release, moving from alpha to closed beta this week. It's a direct signal of how ad planning, buying and optimization get reorganized around agents before rivals lock in the pattern.
Platforms like LTK are rolling out AI dashboards that automate creator discovery, pricing, contracting and measurement, a bid to make influencer marketing as buyable as programmatic display. Holdcos have spent two years snapping up influencer tech for the same reason. But the counter-case is blunt: creator work is "art and science," and the automation that scales it can strip out the authenticity that made it work.
Google is answering the biggest complaint about AI Max, its fastest-growing AI bidding product: that it burns spend with too little control. A year in, it is rolling out AI Brief, a Gemini-powered steering layer where advertisers set messaging, matching and audience boundaries in plain words, plus expanded reach into Shopping and travel. Practitioners who felt the feature was pure negative-keyword overhead finally get an explicit, pre-commit feedback loop in the driver's seat.
Advertisers on ChatGPT complain they can't tell OpenAI who to reach, where ads can run, or where they actually appeared. So OpenAI is quietly testing negative targeting: letting a small group of buyers flag contexts they don't want to appear beside. It's early, details are scant, and there's no timeline. Any fix to that opacity would be overdue.
You’re caught up.
That was all 9 stories in the 27 August issue.