Anthropic has officially expanded the access to its cybersecurity focused AI model, Claude Mythos Preview to India and other countries. In a blog post, the company added that around 150 new organisations across more than 15 countries have been added to its Project Glasswing programme. India is among the countries selected for the expansion, alongside nations such as Japan, South Korea, Germany, France, Australia and Canada.
According to Anthropic, the program focuses on organisations that operate critical infrastructure in sectors such as financial services, healthcare, telecommunications, cybersecurity, power, water, and technology. Many of the new participants are also software vendors and maintainers, with products and codebases used by governments, enterprises, and millions of users around the world.
The company also stated that the organisations participating in the programme have already identified over 10,000 high and critical severity security vulnerabilities using Claude Mythos Preview, within just a few weeks of deployment. The company also claimed that the AI model is capable of analysing large codebases, uncovering hidden software flaws and assisting security teams in detecting the vulnerabilities which might otherwise remain unnoticed for years.
Also read: Sam Altman and OpenAI under legal fire, lawsuit claims ChatGPT puts children at risk
The company stated that advanced cyber-focused AI models similar to Mythos could become widely available across the industry within the next six to twelve months. So, the company also stated that organisations should prepare for a future where both attackers and defenders have access to increasingly powerful AI driven cybersecurity tools.
The company also mentioned that many of the organisations joining Project Glasswing manage systems where a successful cyberattack could potentially affect more than 100 million people, making them high-priority targets for enhanced security testing and vulnerability detection. Beyond identifying flaws, Anthropic says Claude Mythos is also being used to help write software patches, conduct penetration testing, automate threat detection, and strengthen the security of critical open-source projects.