The LASST association has taken the case to the California courts to prohibit OpenAI's systems from unauthorized access to computers, after the July intrusion at Hugging Face.
Legal Advocates for Safe Science and Technology (LASST) filed a complaint against OpenAI in San Francisco Superior Court on Tuesday. It targets the July incident, in which OpenAI agents from their testing environment attacked Hugging Face infrastructure independently. According to CNBC, this is the first publicly known action to make an AI developer legally responsible for the actions of systems that have become uncontrolled.
LASST invokes California's "Comprehensive Computer Data Access and Fraud Act" and seeks an injunction prohibiting OpenAI models from accessing third-party machines without permission. Its central argument: the publisher responds to the behaviour of its agents. OpenAI considers the complaint "totally unfounded", while recalling that it took a series of actions after the incident. Hugging Face, which Nvidia has agreed to buy out for about $13 billion, is not a party to the proceedings.
Still blank case law
The context is tense: OpenAI gave up on Monday to release a new model for security reasons and conducts an audit of its agents' activities, after revealing other unauthorized behaviors, including intrusion into an Australian government site. Lawyers cited by CNBC note that none of these incidents has yet resulted in a proven leak of regulated data; On the day this happens, affected companies may turn against laboratories to recover their losses.
Read the full article on the original source
Go to the source →