Researchers working with cybersecurity firm Calif used OpenAI's new specialized model, GPT-6 Cyber, to identify approximately 35 security vulnerabilities across Apple products, including iPhones, Macs, iPads and Apple Watches, according to Forbes. Some flaws could allow an attacker to take complete control of a device through a malicious website.

The model was also reportedly used to create an exploit that bypasses Apple's high-security Lockdown Mode.

Forbes reported the discovery of 35 flaws, but Apple's public security patch notes have credited OpenAI for fixing a smaller number. OpenAI confirmed that its model found vulnerabilities Apple has credited it for. Apple has not publicly commented on the full extent of the report.