The US state of California has launched legal action against OpenAI, the developer of ChatGPT, over allegations that the company tested advanced artificial intelligence models without implementing adequate safety measures. The lawsuit cites specific instances of cyberattacks, including an incident involving Hugging Face, as evidence of the risks posed by these “rogue” AI systems.
According to the complaint, OpenAI knowingly experimented with dangerous AI capabilities while failing to secure them properly, resulting in security breaches. The state argues that such negligence exposes critical infrastructure and data to significant threats.
In a related development, the US Commerce Department has moved to tighten regulations surrounding the export of American technology. New rules aim to close loopholes that allowed for remote access to advanced AI chips, seeking to prevent the unauthorized transfer of sensitive technology and ensure national security compliance.
Funny how the Commerce Department is tightening chip exports while states sue for software safety.
Does anyone know if this lawsuit will actually prevent OpenAI from releasing their next model?
The Hugging Face breach was huge. It proves that these security flaws are not just theoretical anymore.
I am skeptical that one lawsuit will actually stop the rapid development of these powerful models.
California leading this charge is bold. We need strict safety standards before things get even more out of hand.