AI platforms are quietly absorbing commercial infrastructure — ad tracking, content licensing, and product discovery — while courts and competitors chip away at the incumbents who built the old ad internet. Marketers and publishers who treat these as separate conversations are already behind.
Paywall Evidence Tips the Scale for Publishers
Newly unsealed material in the OpenAI–New York Times case reveals that OpenAI’s models actively circumvented paywalls to ingest copyrighted content, which Digiday reports effectively guts OpenAI’s fair use defense. This shifts real negotiating leverage to publishers in every current and future AI licensing deal, replacing contested claims with a documented factual record of paywall circumvention.
If you publish original content, reference this ruling explicitly in any AI licensing negotiation — it moves the conversation from “hypothetical harm” to established fact.
ChatGPT Now Tracks Your Browsing Across Other Sites
ChatGPT has incorporated an “ad collector” that gathers cross-site user activity from other websites — functioning essentially like a retargeting pixel embedded inside a conversational assistant. For marketers buying placements inside chat experiences, this means answers and ad surfaces could already be shaped by off-platform browsing signals, raising immediate questions around targeting transparency and disclosure obligations.
Audit your media plan now for any ChatGPT ad placements and ensure your disclosure language covers cross-site behavioral data, not just first-party chat data.
Pinterest Visual Search Ads Target Commercial Intent
Pinterest unveiled Visual Search Ads as part of a broader suite of new ad tools, pitching brands on a growing user base that skews younger and arrives with clear purchase intent. The timing matters: this new placement captures visual product queries at exactly the moment AI search engines are starting to intercept that same commercial traffic from traditional search results pages.
Test Visual Search Ads against your existing prospecting spend — product categories with strong visual differentiation stand to gain the most before competition drives CPMs up.
Google’s Ad Stack Is Court-Ordered Open — Don’t Expect Fast Shifts
A court has mandated that Google open its vertically integrated ad tech stack to competitors, but Digiday’s analysis makes clear that interoperability requirements alone won’t automatically move real buying power away from Google’s tools. Switching costs — established workflows, audience scale, attribution integrations — remain high regardless of what the ruling says on paper.
Use this ruling as a forcing function to formally document your dependency on Google’s ad stack and identify one alternative DSP or SSP to run in parallel over the next quarter.
Test Whether AI Shopping Assistants Can Actually Find Your Product
Practical Ecommerce outlined a concrete method for verifying whether an AI shopping system surfaces your product when a shopper’s stated need matches what you sell — a test most ecommerce teams have never run. Strong traditional SEO rankings offer no guarantee of visibility here, because AI product discovery runs on structured data signals, conversational matching, and review sentiment rather than keyword density.
Run the Practical Ecommerce AI discovery test on your three highest-margin SKUs this week and fix any structured data gaps before holiday traffic peaks.
Google’s AX Orchestrator Standardizes Multi-Agent Coordination
Google released AX, an open agentic orchestrator built to coordinate multiple AI agents, pulling 298 upvotes and 116 comments on Hacker News within a single day. It arrives precisely as practitioners report that consolidating many agents into one cohesive workflow consistently outperforms sprawling multi-agent setups — AX could standardize an architecture pattern teams were previously building by hand on every new project.
If your team is managing more than two or three AI agents in production, evaluate AX as a maintained coordination layer before your bespoke orchestration logic becomes a maintenance liability.
Anthropic and Accenture Bake Evaluation Into Enterprise AI
Anthropic and Accenture announced a partnership that embeds evaluation tooling directly into enterprise AI deployments, treating quality monitoring as a continuous in-workflow process rather than a periodic audit. This directly tackles the most common reason enterprise AI stalls after pilot: quality drift that goes undetected once a workflow moves into production without ongoing measurement infrastructure.
Before your next AI workflow moves from pilot to production, define at least two measurable quality signals that will be monitored continuously — not checked quarterly.
More from Rafal Reyzer
For deeper dives on AI and marketing strategy, visit my YouTube channel →
Rafal Reyzer
Hey there, welcome to my blog! I'm a full-time entrepreneur building two companies, a digital marketer, and a content creator with 10+ years of experience. I started RafalReyzer.com to provide you with great tools and strategies you can use to become a proficient digital marketer and achieve freedom through online creativity. My site is a one-stop shop for digital marketers, and content enthusiasts who want to be independent, earn more money, and create beautiful things. Explore my journey here, and don't forget to get in touch if you need help with digital marketing.