High-Level Meeting on AI Model Mythos
Union Finance Minister Nirmala Sitharaman conducted a high-level meeting to assess the risks posed by Anthropic's latest AI model, Mythos, on India's banking sector. Concerns are that the model's capabilities could pose significant risks, leading to preemptive steps for safeguarding systems.
Participants and Concerns
- Attendees included IT Minister Ashwini Vaishnaw, representatives from the Reserve Bank of India, and the banking industry.
- A report revealing unauthorized access to Mythos raised significant concerns.
- Discussions focused on building safety and security around AI models with regulators like the RBI.
Capabilities of Mythos
- Mythos is described as Anthropic's most powerful AI model, capable of autonomously identifying serious vulnerabilities.
- The model scored 73% on expert-level hacking tasks in tests by the UK's AI Security Institute.
- Its unprecedented ability is causing global governments to strengthen cyber defenses.
Government Response
- The Indian government is in talks with Anthropic's US leadership and has contacted the Ministry of External Affairs.
- The IT Ministry established the AI Governance and Economic Group (AIGEG) for AI policy coordination.
- A separate Technology and Policy Expert Committee (TPEC) was formed to provide expertise to the governance group.
- Key information security officers in critical sectors, including banking, have been sensitized to the risks posed by Mythos.
Global and National Implications
With Mythos's capability to expose vulnerabilities in critical systems, global governments are enhancing their cyber defenses. In India, agencies like Cert-In are actively involved in monitoring the situation, addressing unauthorized access concerns, and planning next steps for protecting key sectors.
