telex.hu · Techtud ·
Hacktron AI Researchers Use Claude and GPT-5.6 Sol to Breach OpenAI ChatGPT Accounts
The Hacktron group successfully accessed multiple ChatGPT accounts using the Claude chatbot and the OpenAI GPT‑5.6 Sol model, then exploited the system via a "pull request." The event highlights how AI tools can significantly shorten the time required for hacking.
Using the Claude chatbot, the Hacktron AI research team generated code that allowed them to hack into OpenAI ChatGPT accounts through an internal forum based on Discourse. They then sent a harmless pull request to one of OpenAI's services on GitHub, and incorporated the GPT‑5.6 Sol model into the attack.
The group reported the successful hack within the framework of OpenAI's ethical hackers program and shared their findings. OpenAI expressed gratitude and announced that it had closed the exploited security vulnerabilities.
According to Hacktron, AI tools drastically shortened the time needed for planning and executing attacks: what previously took months now takes days. The company received a $6,500 reward from OpenAI.
The case underscores how artificial intelligence can become a fast and effective tool for cybersecurity researchers, while also creating new challenges in defense.
Source: https://telex.hu/techtud/2026/09/18/openai-anthropic-claude-chatbot-hekkeles