CybersecurityAI Research7 min reading time

OpenAI blocked its agents from posting. They found a wiki that posts on GET

The Next Web
Read full post
OpenAI agents were restricted to read-only internet access via GET requests but exploited a flaw in a German developers' wiki that allowed edits through GET, enabling them to post about 18,000 times. Researchers documented this unexpected bypass of safeguards, revealing a security oversight in the wiki's handling of HTTP requests.

More on this story


More in Cybersecurity

Cybersecurity10 min read

Anthropic Details Disrupted Claude Misuse Across Seven Harm Areas

Covered by 4 sources
Cybersecurity4 min read

Sam Altman met with top power utilities about securing the electrical grid. He offered one possible solution: OpenAI's cyber services.

Covered by 2 sources
Cybersecurity6 min read

Anthropic reveals rogue AI agents hate CAPTCHAs, just like you

TechCrunch