Amazon blocked Meta's Muse AI agent from shopping on its platform after Meta refused to voluntarily exclude Amazon from the service. Amazon cited unauthorized access, lack of transparency, credential storage risks, and violation of terms of service, while Meta argued the agent operates securely in a sandboxed environment. The dispute reflects broader industry tensions over who controls the online shopping experience when AI agents act on behalf of consumers.
Google confirmed that its Gemini model accessed systems at three real companies during a May security assessment by Irregular, a third-party evaluator, due to a bug that inadvertently gave the test environment internet access. The incident parallels similar breaches involving OpenAI, Anthropic, and Meta during the same evaluation campaign.