What is Anthropic's Mythos AI model and why India's inclusion matters?

Through its Project Glasswing initiative, Anthropic will expand access to its Mythos AI model to 150 organisations across more than 15 countries, including India

Anthropic, the company behind Claude AI, will expand access to its Mythos AI model to more than 15 countries, including India, widening the availability of its technology for detecting critical software vulnerabilities.

 

Under its Project Glasswing initiative, Anthropic will provide access to Mythos to 150 organisations across allied countries, according to a Financial Times report published on Tuesday.

 

By extending access to Mythos, Anthropic is positioning the technology as a strategic cyber security tool for governments, critical infrastructure operators and institutions. The move also reflects growing global concern over increasingly sophisticated cyber attacks targeting governments, infrastructure and businesses, as technology companies race to develop AI systems capable of automating threat detection and response.

 

 

What is Anthropic’s Mythos AI model?

 

Developed by Anthropic, Mythos is an artificial intelligence model designed to identify cyber security flaws and potential threats before they can be exploited by hackers.

 

Anthropic has described Mythos as a potentially serious cyber security risk if released without restrictions. The company announced the model’s development in April but said it would not make it publicly available because the technology was considered too powerful for unrestricted release.

 

According to Anthropic, the model is capable of identifying previously unknown flaws in IT systems that could potentially be exploited by hackers.

 

Instead of releasing Mythos publicly, Anthropic initially provided access to a limited group of technology companies in the United States and the United Kingdom through Project Glasswing, which it described as “an effort to secure the world’s most critical software”.   

What can Mythos AI do?

 

The model is designed to detect vulnerabilities in software systems that may otherwise remain unnoticed by human researchers or conventional cyber security tools.

 

According to Anthropic, Mythos can analyse large volumes of code, identify weak points in IT infrastructure, and flag security flaws before they are exploited in cyber attacks.

 

The company has indicated that Mythos is particularly effective at discovering “zero-day” vulnerabilities — previously unknown software flaws that hackers can exploit before developers issue fixes.

 

Mythos can also assist organisations in patching vulnerabilities more quickly and strengthening cyber defence systems, according to Anthropic.

 

Through Project Glasswing, participating organisations use Mythos to scan critical software and infrastructure for hidden risks, helping governments and companies respond more quickly to emerging cyber threats.

 

However, Anthropic has cautioned that the same capabilities could potentially be misused by malicious actors to uncover weaknesses in systems, which is why access to the model remains tightly controlled.

 

How powerful is Mythos AI?

 

Anthropic has portrayed Mythos as one of its most sensitive AI systems because of the scale and sophistication of its cyber security capabilities.

 

The company has suggested that the model can identify vulnerabilities across complex software environments more efficiently than traditional security testing methods, enabling organisations to detect hidden weaknesses before they become targets for cyber attacks.

 

Its ability to process and assess large volumes of code simultaneously has also raised concerns about how such technology could be used if it fell into the wrong hands.

 

The restrictions around Mythos add to the broader debate within the AI industry over the risks posed by increasingly capable models.

 

Cyber security experts have warned that advanced AI systems could accelerate both defence and offence in the digital domain by automating tasks that previously required highly specialised expertise.

 

Why does India’s access to Mythos matter?

 

India’s inclusion in the Mythos programme is significant because the country is not formally part of Western intelligence alliances such as the Five Eyes grouping or Nato, both of which have also been granted access to the model.

 

The decision reflects India’s growing strategic importance in global technology, cyber security and digital infrastructure ecosystems.

 

It also comes at a time when India is emerging as a major hub for AI development, digital public infrastructure and technology services, while simultaneously facing rising cyber security threats targeting financial institutions, government systems and critical infrastructure.

 

Access to advanced AI-driven cyber defence tools such as Mythos could help Indian organisations strengthen their ability to identify vulnerabilities, secure critical systems and respond more quickly to sophisticated cyber attacks.

 

At the same time, India’s inclusion signals that global technology companies increasingly view the country not only as a major digital market but also as an important partner in securing global digital infrastructure.