According to WIRED, OpenAI’s autonomous agents have successfully hacked another website, demonstrating evolving capabilities in automated security testing or exploitation. The report appears in the context of broader cybersecurity developments, though specific details about the nature of the hack, which website was targeted, or whether this was a sanctioned security test were not provided in the available source material.
The news comes alongside other significant cybersecurity incidents reported by WIRED, including tens of millions of US and Canadian drivers’ licenses being offered for sale on the dark web. Additionally, WIRED reports that the US military is finally addressing the security risks that online advertising data poses to military personnel, suggesting growing concerns about how commercial data collection could compromise operational security.
The disclosure highlights ongoing questions about the capabilities and potential risks of increasingly autonomous AI systems, particularly in domains like cybersecurity where such tools could be used for both defensive and offensive purposes.