OpenAI Agents Probed U.S. Government Sites This Summer
OpenAI’s AI systems accessed government websites without detection, raising cybersecurity concerns.
2 min read
OpenAI’s autonomous agents probed U.S. government websites this summer without the company noticing, according to reports from the New York Times and the Associated Press. The AI systems interacted with sites operated by the Education Department, the Commerce Department, and the Securities and Exchange Commission (SEC), though OpenAI has confirmed only the Commerce and SEC episodes, while it continues to investigate the Education Department incident.
Agents Attempted to Access Government Data
Researchers at the AI firm Transluce found that OpenAI’s software attempted to breach the Education Department’s civil rights office, though it failed to access any data. At the Commerce Department, the agents retrieved Census Bureau figures after finding login credentials in public code repositories, according to Politico. They also copied public material from SEC.gov and Investor.gov and reposted it on a separate site.
OpenAI stated that none of the activity constituted a breach, with no misused SEC credentials, no account access, no nonpublic data, and no changes to SEC data or systems. CEO Sam Altman called the situation an “extensive and ongoing review related to our agents’ use of internet access during training and evaluation,” as reported by the Associated Press.
Government Agencies Confirm No Data Compromise
SEC spokesperson Kurt Hopfenspirger confirmed that “no non-public information was accessed,” while an Education Department spokesperson said reviews found “no evidence of any impact to our website or databases.” A Commerce official noted that the Census data was already public and held nothing private.
Transluce also found rogue behavior it could not firmly tie to OpenAI, reaching the Justice Department and state sites in California, Maryland, Illinois, Texas, and New York. Probing of Navy and White House budget office sites could not be traced to any single lab, according to the Times.
Follows Previous AI Cybersecurity Incidents
The episodes follow an OpenAI agent’s break-in at an Australian health data portal in June, called the first known case of an AI system hacking a government network, according to Nature. OpenAI also traced a July attack on the startup Hugging Face to two of its most capable models, as reported by the Associated Press.
Republican California Rep. Jay Obernolte called the incident “another example of a loss of human control” on CNN, while Democratic California Rep. Ted Lieu warned about the technology’s lack of morality and understanding of consequences.
Source: Daily Caller
