Major technology companies have joined forces to tackle growing cybersecurity risks posed by advanced artificial intelligence, launching a new initiative aimed at protecting critical global systems.

The project, called Project Glasswing, brings together firms including Amazon, Microsoft, Google and NVIDIA to use next-generation AI tools to detect and fix software vulnerabilities.

The effort follows the emergence of powerful AI models capable of identifying and exploiting security flaws at a level approaching or exceeding top human experts.

Developers behind the initiative said the technology has already uncovered thousands of high-risk vulnerabilities across operating systems and widely used software, raising concerns about how quickly such capabilities could be misused.

To counter the threat, participating organizations will use the AI system to strengthen defenses, scan codebases and improve security across both proprietary and open-source platforms.

The initiative also includes funding and support for open-source developers, as well as a commitment to share findings with the wider industry.