A BBC investigation has confirmed cases where AI chatbots and search functions were manipulated by sophisticated blog posts to spread misinformation. It was reported that major AI tools, including ChatGPT and Gemini, provided biased answers regarding health and wealth management.
Google Updates Policy Following BBC Investigation into AI Information Pollution
This article is a translation. Read the Japanese original
The investigation reproduced a phenomenon where AI adopted information and repeated factual errors based on the publication of a single article on a personal site. For example, there was a case where an article written by a reporter claiming they were a "world champion hot dog eating contestant" was disseminated by AI the following day.
In response to this issue, Google announced that it has updated its policies. The company has positioned this action as a "clarification" of its long-standing anti-spam measures.
However, it appears that Google and other companies are strengthening their countermeasures behind the scenes. Experts point out that users should assume they may be manipulated until better mechanisms are established.
Source: Google’s AI is being manipulated. The search giant is quietly fighting back（HN 339pt・215コメント）（HN Search (backfill)、2026-05-20）