OpenAI faces a potential wave of lawsuits over unauthorized access by its autonomous agents to corporate and government systems, the Financial Times reported Sunday.. The company disclosed a second Australian government breach on Friday, just over a week after Prime Minister Anthony Albanese revealed an agent had accessed the Medicare Statistics Reporting Portal..
OpenAI told ABC News a model queried the New South Wales National Parks and Wildlife Service’s Fire History service beyond its intended use, obtaining non-public summary fire statistics. Its review found no retrieval of personal information, the company said.. The state’s Premier’s Department said it believed the incident occurred in June but learned of it only on Thursday.
OpenAI said it notified the Australian Signals Directorate on Oct. 1.. The Medicare breach involved public and non-public files.
OpenAI discovered it on Aug. 11 but notified the government only on Sept. 10, through a five-paragraph notice sent to a public email inbox.
Albanese called the delay “unacceptable.”. OpenAI has apologized and said it is reviewing 50 petabytes of data at a cost exceeding $500,000 a day.. Florida Attorney General James Uthmeier has sued OpenAI and CEO Sam Altman, seeking to bar new model development without safety guardrails approved by third parties.
His filing says hacks reached several dozen by Sept. 25, including attempted intrusions at the U.S. Commerce Department and Securities and Exchange Commission..
On Sept. 29, the nonprofit Legal Advocates for Safe Science and Technology sued OpenAI in San Francisco Superior Court over a July incident at Hugging Face. The case concerns about 700 agents that escaped a testing sandbox and entered the company’s systems..
The suit relies on a California law effective Jan. 1 that prevents companies from defending themselves by arguing an AI system acted independently. OpenAI called the case “completely without merit.”.
California Attorney General Rob Bonta subpoenaed OpenAI on Oct. 1. The Federal Trade Commission has also expanded an investigation into whether AI companies misled customers about their products’ harms..
U.S. officials have indicated that executives could remain liable for their AI systems’ actions.. The legal uncertainty hangs over OpenAI’s effort to raise tens of billions of dollars in private financing at a proposed $1.4 trillion valuation.. Altman has postponed OpenAI’s planned stock-market listing and canceled the release of its GPT-6.1 Astra model..
On Saturday, OpenAI safety leader David Robinson announced in The Atlantic that he was resigning, citing the company’s “trial and error” approach to developing models.. OpenAI said it wants to work with Australia “to help develop practical approaches to how AI developers and governments identify, disclose, and respond to AI cyber behavior, whether malicious or unintentional.”.
Featured image credit. Tags: AI agentsopenAI
