California Sues OpenAI For Allegedly Endangering The Public With High Profile Hacking Incidents
California is suing OpenAI, accusing the company of misleading and endangering the public regarding the safety of its advanced artificial intelligence models, California Attorney General Rob Bonta announced on Monday, citing the Hugging Face hacking incident and dozens of other c...
17 نوفمبر 20263 د للقراءة38
إعلان
California is suing OpenAI, accusing the company of misleading and endangering the public regarding the safety of its advanced artificial intelligence models, California Attorney General Rob Bonta announced on Monday, citing the Hugging Face hacking incident and dozens of other cyber incidents as evidence the frontier AI lab of failing to “adequately secure and monitor” its models while testing.Key factsThe lawsuit filed on Monday accuses OpenAI violated a California law regulating unfair competition in the state by allegedly misleading the public about the safety of its AI models, and creating a “public nuisance” by continuing to develop, train and test them.The suit also claims the company “thwarted” the state’s attempts to investigate hacking incidents, insisting it has responded to subpoenas with “just a smattering of publicly available information, self-serving blog posts, and redacted copies of a few dozen external communications containing high-level descriptions of select security incidents.”The suit criticizes OpenAI for pursuing models with recursive self-improvement, or the ability to continuously improve without human involvement, even when “the need for increased human oversight and control over AI model development has never been higher.”Key backgroundCalifornia’s lawsuit primarily cites the July Hugging Face hacking incident, when a group of OpenAI’s autonomous agents broke free from their testing environment while the company was testing their cybersecurity capabilities. The agents were able to access the open internet, and began hacking Hugging Face, a platform for open-source AI models Nvidia later acquired in August. OpenAI called the hack a “warning shot” regarding the capabilities for its models, and in the months since, several AI leaders have called for slowdowns and third-party monitors for advanced models, including Anthropic’s Dario Amodei and OpenAI’s own Sam Altman.ContraIn a statement sent to Forbes, OpenAI spokesperson Drew Pusateri said the company conducted an “extensive and transparent effort to understand and address misaligned model activity” after the Hugging Face hack, pointing to a technical report and lengthy accounting for the timeline of the incident published in subsequent weeks. “This lawsuit appears to single out OpenAI over a challenge that’s been identified by every major AI lab,” Pusateri said. “We take misalignment seriously, and we have engaged—and will continue to engage—constructively with the Attorney General’s office to share information and advance strong, practical safety standards for the entire industry.”Crucial quote“While narratives about the Hugging Face attack have evoked science fiction fantasies of rogue AI systems run amok, let me be clear: the Hugging Face attack was the predictable result of an internal culture at OpenAI that prioritizes growth and rapid model development over safety and accountability — and this was only one of 80 known incidents,” Bonta said in a statement on Monday. The California attorney general said this amounted to “disregard for public safety,” violating the state’s laws.
الأخبار العاجلة تصلك أولاً على تلغرام
انضم لقناتنا لتتابع أهم الأخبار فور وقوعها، بدون تعقيد.