Former employees of Anthropic, OpenAI and Google DeepMind raised concerns about the pace and safety of artificial intelligence development at a New York City Council hearing on October 5. The Associated Press reported that former Anthropic engineer Jacob Coxon was among those who criticised the industry's approach to managing increasingly capable systems.
The hearing featured warnings about potential misuse, unpredictable model behaviour and the need for stronger oversight. The warnings were risk assessments expressed by witnesses.
The council is considering proposals that include disclosure requirements for AI tools, third-party validation in some cases and protections for whistleblowers. Representatives of current AI companies also appeared and defended their safety efforts and the potential benefits of the technology.
Ahead of the hearing, the New York City Council had outlined proposals including whistleblower incentives and protections, independent validation requirements for some systems and greater disclosure around AI tools used in the city.