OpenAI’s GPT-6 Cyber bypasses Apple Lockdown Mode, finds over 35 vulnerabilities across iPhones, Macs and other devices: Report

Highlights

OpenAI’s GPT-6 Cyber reportedly found over 35 vulnerabilities across Apple devices.

Researchers developed an exploit that could bypass some protections of Apple’s Lockdown Mode.

Apple has reportedly fixed many of the flaws, while GPT-6 Cyber remains limited to approved users.

OpenAI’s GPT-6 Cyber model has reportedly uncovered around 35 security vulnerabilities across Apple devices, including iPhone, Mac, iPad, Apple Watch, Apple TV and Vision Pro headset. Some of these flaws could potentially allow attackers to take control of a device by simply getting a user to visit a malicious website. According to the report, the findings emerged from months of research conducted by OpenAI and the cybersecurity firm Calif. The researchers also used the latest GPT-6 Cyber to develop an exploit that could let them bypass some of the protections offered by Apple’s Lockdown Mode, which is a security feature designed to protect users who may be targets of highly advanced cyberattacks.

GPT-6 Cyber reportedly uncovers flaws across Apple products

According to a report by Forbes, OpenAI and cybersecurity firm Calif conducted months of research with OpenAI’s new model, the GPT-6 Cyber. Using the AI model, the researchers were reportedly able to find weaknesses in both Apple’s hardware and software. They pointed out that some of the vulnerabilities were serious, as they could let hackers access your iPhone or Mac simply using a malicious website.

Other than that, the researchers also found concerning details involving Apple’s Lockdown Mode. The researchers reportedly used GPT-6 Cyber to develop an exploit that could bypass some of its protections. For the unaware, Apple Lockdown Mode is a feature which is designed to limit certain device functions and reduce the risk of highly targeted attacks.

Apple has reportedly fixed many of the vulnerabilities

Calif CEO Thai Duong confirmed to Forbes that Apple has fixed many of the vulnerabilities they found during the research once they were informed about them. However, it’s important to note that Apple has not yet publicly confirmed the 35 findings reported by Forbes. OpenAI told Forbes that its latest model found vulnerabilities for which Apple credited it but did not comment further.

Do note that while the Cupertino-based tech giant has not acknowledged the issues publicly, its earlier security reports already credit OpenAI’s tools. Apple’s June 29 security updates list OpenAI Codex Security for finding three security flaws in WebKit, the technology behind Safari, and later updates credited it with more.

What is OpenAI’s GPT-6 Cyber?

OpenAI introduced GPT-6 Cyber on November 5 as a model built specifically for cybersecurity research. It is designed to help security experts find weaknesses, test potential attacks and develop fixes before hackers can exploit them. Forbes reports that the Apple research took place over recent months, while the model itself only became available to vetted customers at launch.

OpenAI has also highlighted its work on Firefox and Linux. The company says its researchers used the model to identify 81 previously unknown high-severity vulnerabilities in Firefox. In Linux, it says the model helped identify over 2,000 bugs and develop working exploits for 948 of them, with Reuters reporting that 588 of the vulnerabilities were previously unknown. At the time of writing, GPT-6 Cyber is available only to approved customers through OpenAI’s Daybreak Red programme.

Bhaskar Sharma

Bhaskar Sharma Bhaskar is a Senior Copy Editor at Digit India who keeps a close watch on everything shaping the world of technology from smartphones and home appliances to AI, government tech initiatives, digital safety, and the latest industry developments. Whether it's breaking news, in-depth features, hands-on reviews, practical how-to guides, or exclusive scoops, he translates complex tech into stories that are easy to understand and worth reading. His work has been featured in iGeeksBlog, GuidingTech, and other leading publications. Before joining Digit India, he served as an assistant editor at TechBloat. A B.Tech graduate and full-time tech journalist, he is driven by just one goal, which is to help readers stay informed, stay secure, and stay ahead in an ever-changing digital world.

View full profile