LOADING...
Indian government agencies, select private firms gain access to Mythos
Mythos can identify software vulnerabilities before they are exploited

Indian government agencies, select private firms gain access to Mythos

Jun 07, 2026
03:18 pm

What's the story

In a major development, Indian government agencies and select private sector firms have been given access to Anthropic's advanced artificial intelligence (AI) model, Mythos. The move comes as part of the global expansion of Project Glasswing. The number of Indian entities with access is still in single digits, sources familiar with the matter told PTI.

Cybersecurity initiative

Mythos can identify software vulnerabilities before they are exploited

Project Glasswing, which Anthropic describes as a "collaborative effort to secure the world's most important software," aims to improve cybersecurity by giving select organizations access to Claude Mythos Preview. This advanced AI model can detect software vulnerabilities before they're exploited. When first announced in April, around 50 partner organizations were given access, collectively identifying over 10,000 high or critical-severity security flaws across various codebases.

Program growth

New partners include vendors maintaining codebases relied upon globally

Earlier this week, Anthropic announced a major expansion of Project Glasswing, adding some 150 more organizations across over 15 countries. The new cohort includes operators of critical infrastructure in sectors such as power, water, healthcare, communications, and hardware. Many of these new partners are vendors that maintain codebases relied upon by numerous other organizations worldwide, including governments.

Advertisement

AI caution

Anthropic calls for pause in AI development amid risks

Along with the expansion of its cybersecurity initiative, Anthropic has also raised alarms over the rapid pace of artificial intelligence (AI) development. In a recent blog post titled "When AI builds itself," the company suggested that leading artificial intelligence labs should keep the option to slow down or even pause frontier AI development on the table. The firm warned that increasingly capable AI systems may one day gain the ability to accelerate their own development beyond human oversight.

Advertisement