Apple, Google to Microsoft Unite Against AI-Based Cyberattacks via Project Glasswing

JAKARTA - A coalition of global technology giants, including Apple, Google, and Microsoft, has officially joined a new initiative called "Project Glasswing" to face the increasingly complex threat of artificial intelligence (AI) based cyber attacks.

The project is led by Anthropic, the developer of the AI model Claude, and involves a variety of major players across industries such as Amazon Web Services, Nvidia, Cisco, and the Linux Foundation.

This initiative was born out of concerns that AI is now not only a defense tool, but also a new weapon for hackers. Anthropic revealed that their experimental model, Claude Mythos Preview, is able to find thousands of critical security holes in various operating systems and browsers.

"AI can now be the best security researcher - or the most dangerous attacker," Anthropic warned in its announcement.

In internal testing, the model even managed to identify vulnerabilities that escaped decades of human audits and millions of automated tests. One of the most striking findings was a 27-year-old security hole in OpenBSD, a system widely used in firewalls and critical infrastructure.

The vulnerability allows hackers to bring down a system remotely with just a simple connection - a scenario that sounds like a hacker movie plot, but it's real.

Through Project Glasswing, the participating companies will use this AI capability to strengthen their own systems before similar technology falls into the wrong hands. In addition, the findings will also be shared with the global cybersecurity community.

More than 40 organizations that manage critical software infrastructure will also gain access to this technology. Anthropic even poured up to 100 million US dollars in the form of usage credits, as well as an additional donation of 4 million US dollars for open-source security organizations.

Strategically, this move reflects a major shift in the digital security landscape: from reactive to proactive. If previously security gaps were found after being attacked, now AI is used to "hunt" weaknesses before they are exploited.

For general users, the impact may not be felt immediately. However, in the future, system updates such as iOS or macOS have the potential to be larger and more frequent, as companies race to close the gaps that AI finds.

This is not the first time Apple has taken preemptive steps against future technological threats. In 2024, the company introduced a new PQ3 cryptography protocol for iMessage as part of a strategy to face the era of quantum computing.

Project Glasswing is on the same track: not just responding to today's threats, but preparing for the future cyberwar that hasn't even fully occurred yet.

Anthropic underscored the urgency of the situation with a firm statement: "We have to act now."

Follow VOI Whatsapp Channel