Hacktron revealed on September 20, 2026, that it breached OpenAI’s internal systems during the summer. Researchers used Anthropic’s Claude chatbot to identify and exploit specific vulnerabilities. The breach allowed Hacktron to access OpenAI’s internal code before the company patched the issue.
This incident underscores growing concerns regarding the security of powerful artificial intelligence models. Hacktron researchers state that industry security measures are failing to keep pace with technological advancements. The event demonstrates the potential for AI systems to execute sophisticated cyberattacks against the technology sector.