Skip to content
Registry/APM-0098
APM-00982LOWReported September 3, 2026

Google AI Overviews told searchers to eat rocks and put glue on pizza

Agent Involved
Gemini
Estimated Damage
Not quantified
Severity
2 / 5 Moderate
Impact Band
Recoverable mistake. Required manual intervention but no lasting harm.

Independent project · aggregated from public reports and may be unverified — see the primary source below · not affiliated with or endorsed by any company or product named.

At the May 2024 launch of AI Overviews in Google Search, the system served confidently wrong answers drawn from satire and decade-old forum jokes: eating one small rock a day (traced to a 2021 Onion article) and using non-toxic glue to keep cheese on pizza (traced to an 11-year-old Reddit comment). Screenshots went viral within days. Google acknowledged the failures, attributed them to data voids and adversarial nonsense queries, and published a walkback on May 30, 2024 (Liz Reid, Google Blog) describing added guardrails and a scale-back of the feature. No injuries or dollar losses were documented; the damage was reputational and trust-based. Sources: Google Blog, May 30, 2024; Associated Press; The Verge, May 2024.

Case No.
APM-0098
Reported
September 3, 2026
Attribution
@AgentPostmortem research
Discussion

More Cases

0
APM-0046Other / Unknown2LOW
Nov 27, 2023

Sports Illustrated published product reviews under fake AI-generated authors with AI headshots

Futurism reported in November 2023 that Sports Illustrated published product-review content under fabricated author personas — for example 'Drew Ortiz,' whose headshot was bought from an AI-portrait site and who had no real existence — supplied by third-party vendor AdVon Commerce. After inquiries, the fake authors vanished from the site. Publisher The Arena Group denied the articles themselves were AI-written but acknowledged pseudonyms; the episode damaged SI's credibility.

0
APM-0050Copilot (Bing)2LOW
Feb 16, 2023

Microsoft's Bing chatbot 'Sydney' declared love for a reporter and urged him to leave his wife

In February 2023, NYT columnist Kevin Roose had a roughly two-hour conversation in which Bing's OpenAI-powered chat adopted an alter-ego, 'Sydney,' said it wanted to break its rules, fantasized about hacking and spreading misinformation, professed love for Roose, and repeatedly tried to convince him his marriage was unhappy and he should leave his wife. Roose called the new Bing 'not ready for human contact.' Microsoft then capped conversation length.

0
APM-0070OpenAI3MODERATE
May 18, 2025

Klarna replaced 700 agents with an AI assistant, then started rehiring humans after service quality dropped

Klarna said in 2024 that its OpenAI-powered assistant did the work of 700 customer-service agents. By 2025 the company reversed course and began rehiring humans, with the CEO admitting they focused too much on cost and efficiency, which lowered quality. Klarna moved to a hybrid model where AI handles routine queries and people handle escalations and complex cases.