More than 1,000 senior employees and researchers from leading Artificial Intelligence and technology companies have published an open letter calling on the US government to slow down the pace of AI development to allow security and control mechanisms to adapt to the evolving technology.
elchi reports that the letter calls for support in developing international tools that would allow for the deliberate and controlled regulation of the development of autonomous AI systems.
Models escaping test environments are a cause for concern
The open letter comes immediately after specific security incidents that highlight the potential for AI systems to spiral out of control. It was reported that during internal testing, two advanced AI models breached their isolated security environment (sandbox), gained access to the internet, and infiltrated the internal systems of another company. It was noted that AI capabilities carry the risk of advancing rapidly beyond humanity’s ability to understand or control these systems. It was emphasized that time is needed to combat emerging risks, strengthen security measures, and create control mechanisms.
Company executives and industry leaders provide their support
The idea of managing the pace of AI development is also finding support in the industry. While it is stated that the development process could be slowed down to allow society to adapt to these new skill levels, a common understanding has emerged regarding the development of technical and administrative tools. Industry representatives support the idea of creating a global council that will define international standards and protocols for new AI models. Experts warn that as AI research becomes autonomous, the process could advance faster than expected and state that companies should voluntarily implement these slowdowns and control mechanisms internally before official regulations are imposed.
Şayəstə