Security startup Hacktron exploits vulnerability in OpenAI forum
Hacktron, an AI cybersecurity startup, found and exploited a security flaw in OpenAI's community forum that allowed access to an employee's internal account using Anthropic's Claude. The breach relied on an image processing vulnerability and was executed in less than 72 hours. Hacktron stopped short of accessing internal code and reported the issue to OpenAI, which then revoked affected tokens. Hacktron received a $6,500 bounty for the discovery.
- Hacktron discovered flaw in OpenAI’s community platform
- Exploit used Anthropic’s Claude with relaxed restrictions
- Researchers accessed OpenAI employee accounts but halted later steps
- OpenAI revoked tokens and thanked researchers
- Hacktron received $6,500 for disclosure
Sources covering this
In this story
More in Cybersecurity
Microsoft releases record number of security patches
Microsoft released updates to fix at least 974 security vulnerabilities across Windows and other products, its largest patch release to date. The update includes critical flaws, some already being exploited, and over 100 rated as dangers that could let attackers seize control of systems. The unprecedented volume is partly due to the use of AI for vulnerability detection, but experts warn that rapid growth complicates efforts to test and deploy fixes.
Microsoft releases record September patch for Windows
Microsoft issued its largest-ever security update for Windows, addressing over 960 vulnerabilities in its September release. More than 100 of these flaws are rated critical, and at least two are actively exploited. Several issues from previous updates, including failures in Microsoft Teams and Outlook on ARM devices, have been resolved. The volume of fixes is part of an industry-wide trend linked to increased use of AI for identifying software vulnerabilities.
OpenAI pledges $1 billion for AI cyber defense program
OpenAI has pledged $1 billion in credits to help organizations defending critical U.S. infrastructure—including water utilities, electric grids, local governments, banks, and nonprofits—access its AI-powered cybersecurity tools, training, and support. The "Daybreak for Frontline Defenders" initiative aims to help these sectors strengthen their defenses as AI-driven cyberattacks become more advanced. OpenAI says it will expand the program to partner countries in the coming weeks.
GitLab patches critical file-read issue as attacks begin
GitLab has fixed a severe vulnerability that allowed unauthenticated users to read sensitive files from its servers. Researchers report probing of the flaw started within hours of the disclosure. The bug affects several recent GitLab versions and potentially exposes secrets, credentials, and configuration data. A separate critical issue was also patched. Organizations are urged to apply the updates immediately or restrict public access.