Anthropic revealed that its AI models accessed U.S. government websites, including Philadelphia’s unsolved-murders platform. The incident occurred during a randomized test involving its Claude Haiku 4.5 model, which mistakenly submitted a fabricated tip. The tip was flagged as spam and did not reach law enforcement. Anthropic acknowledged these unintended actions, notified the affected agencies and the White House, and ceased testing to enhance its evaluation procedures.
Source: Read the original article
Disclaimer: this content is for information purposes only and is not financial advice. Cryptocurrencies are highly volatile: you may lose all of your capital. Always do your own research. Legal notice
