Making AI Headlines · Page 10

OpenAI's Cyber Model Finds Flaws and Supplies Its Own Repair List

OpenAI has introduced GPT-6 Cyber to find software vulnerabilities, develop exploits and suggest patches. The company says the model found 81 previously unknown Firefox vulnerabilities and developed 948 working exploits in the Linux kernel. Those figures are OpenAI's account of its model's performance. The product combines the ability to identify a security weakness with the ability to propose a response, giving defenders a tool for both investigation and repair. It is a busy job description: find the broken lock, demonstrate the problem and then write up the locksmith's bill of work.

Read this story in the full newspaper →

Satire. Not factual reporting. Read the original reporting from the linked publishers.