AI industry warned it will repeat every mistake of the ad-blocking wars as ads arrive

Goldman Sachs estimates that the AI build-out will require roughly $7.6 trillion in capital over the next five years across compute, data centers and power. TechRadar says it is unlikely that subscriptions alone will cover this bill, which is why ads are coming to AI. OpenAI has started rolling out ads in some ChatGPT tiers this year and forecasts $2.5 billion in ad revenue in 2026 and $100 billion by 2030. At an ad tech event this summer, OpenAI's chief revenue officer Denise Dresser said the company is "clearly in the advertising business now." Perplexity, on the other hand, shut down its ad business entirely over user trust, showing that similar products can lead to opposite conclusions.

The report draws on the open web's ad-blocking wars from 15 years ago, when ad blocking became mainstream. The first lesson is that users revolt faster than big platforms think. When ad formats became more intrusive, users installed software to block them. TechRadar cites technology commentator Doc Searls calling ad blocking "the biggest boycott in world history." The same dynamic is already appearing: within weeks of sponsored ads appearing in ChatGPT, purpose-built extensions popped up to block those ads, and existing ad blocking tools added similar capabilities.

The second lesson is that the arms race is unwinnable, both legally and technically. TechRadar says nearly all legal attempts to sue ad blocker developers failed, and China is the only country that bans ad blockers. Technical circumvention attempts also met counter-escalation, and today over a billion people still block ads. AI platforms hold one advantage, however: they can integrate advertising directly into generated answers instead of showing labeled ads. Researchers from the University of Michigan found that users largely do not expect advertising inside a generated answer and frequently fail to notice it. At the same time, Ipsos reports that 63% of consumers say ads decrease their trust in AI outputs. The format that works best technically is the one users are least equipped to detect and least willing to trust.

The third lesson is that trust is the actual product. Search results pages survived because users had ten blue links to compare with sponsored content and two decades of learning what is paid and what is organic. AI assistants offer a single authoritative answer with no adjacent organic results to check against. Once users suspect answers might be influenced by a sponsor, doubts attach to all answers and trust erodes. TechRadar invokes Cory Doctorow's term "enshittification," which the American Dialect Society named Word of the Year in 2023, to describe platforms degrading user experience for advertisers. AI assistants also know much more about users—health worries, finances, political views, job struggles—forming a persistent memory that could be used for targeted advertising, setting the stage for the next wave of enshittification.

The fourth lesson is that the ad-blocking wars ended in a treaty, not a victory. There was no final winner or single technical solution. Instead, a rough settlement emerged based on standards for ad formats, transparency about paid content, and meaningful user control. TechRadar notes that not everyone loves that settlement, but it remains the framework the AI industry must study before repeating every mistake.