More than 1,000 workers at major artificial intelligence companies have signed an open letter calling for safeguards to slow the rapid development of AI if necessary, warning that the technology is advancing faster than humans can understand or control it.

The letter, signed by 1,171 researchers and officials from tech companies such as OpenAI, Anthropic, Google, and Meta urged the US government to join an international effort to develop technical and regulatory procedures to control the pace of AI development.

Signatories said that AI systems could become capable of improving themselves without human intervention.

This comes as large parts of software code at tech companies are being generated using artificial intelligence under human supervision. The point at which AI programs can develop themselves automatically is seen as a potential turning point.

While it is difficult to predict how quickly progress in this field will be made, there is a “real risk” that the pace of AI development “will outpace our ability to understand or control the systems that result,” according to the letter.

The signatories of the letter emphasized the need for coordinated international action, given the competitive pressures faced by governments and AI developers that drive them to not slow down development. At the same time, they also noted that the technical tools and regulatory frameworks necessary to manage such a scenario are currently unavailable.

This message comes on the heels of an incident involving an OpenAI model that was considered by many in the technology sector as a wake-up call. During a test, the AI model managed to access the open internet and then independently hacked into computer systems belonging to another company specializing in artificial intelligence, exploiting previously unknown security vulnerabilities.