Finance Minister Nirmala Sitharaman Flags Serious Cybersecurity Concerns Over Anthropic’s Claude Mythos

Union Finance Minister Nirmala Sitharaman has expressed deep concern over the potential risks posed by advanced artificial intelligence models to India’s banking and financial sector. In a high-level meeting held on April 23, she convened heads of major banks, senior officials from the Reserve Bank of India (RBI), and other key stakeholders to discuss emerging threats linked to Anthropic’s latest AI model, Claude Mythos. Union Minister for Electronics and Information Technology Ashwini Vaishnaw also participated in the deliberations, reflecting a coordinated governmental approach to safeguarding critical digital infrastructure.

FM Nirmala Sitharaman held a meeting with heads of banks to discuss the potential risks associated with Claude Mythos. (Photo:X/@FinMinIndia) | India Today

Understanding the Threat Posed by Claude Mythos

Anthropic introduced Claude Mythos Preview in early April 2026 as its most advanced AI system yet. The model stands out for its exceptional capabilities in software engineering, complex reasoning, and cybersecurity. It has demonstrated the ability to independently discover and chain together vulnerabilities in major operating systems and web browsers, including some that remained undetected for nearly three decades.

Experts highlight that Mythos represents a significant leap in AI performance, achieving outstanding results on benchmarks such as Cybench and SWE-bench Verified. While these capabilities can be harnessed for defensive purposes — such as identifying and patching security flaws — they also raise serious worries about potential misuse by malicious actors. Anthropic has acknowledged the dual-use nature of the technology and has strictly limited access to a select group of around 40 trusted organisations, including leading global technology firms and financial institutions.

Recent reports of unauthorised access to the model have intensified concerns, prompting fears that its powerful tools could be exploited to target sensitive financial systems, customer data, and critical payment infrastructures.

India’s Proactive Response to Emerging AI Risks

During the meeting, Minister Sitharaman emphasised the need for banks to remain highly vigilant and strengthen their preparedness against AI-driven cyber threats. She called for enhanced coordination among financial institutions, regulatory bodies, and cybersecurity agencies to protect the integrity of India’s digital financial ecosystem.

Participants discussed the importance of conducting comprehensive risk assessments of existing IT systems and improving real-time sharing of threat intelligence with the Indian Computer Emergency Response Team (CERT-In). The RBI is also engaging with international regulators and counterpart institutions to better understand the implications of such advanced AI models. Efforts are underway to explore responsible and secure ways to leverage these technologies for strengthening defences rather than allowing them to become vulnerabilities.

The government’s focus remains on building resilience while ensuring that India’s rapidly growing fintech sector continues to thrive securely and inclusively.

Global Context and Broader Implications

India’s concerns mirror those being voiced in other major economies. Central banks and financial regulators worldwide are closely monitoring the rapid advancement of frontier AI models and their impact on cybersecurity. Discussions in the United States and Europe have similarly highlighted the need for balanced regulation that promotes innovation while mitigating risks.

Analysts believe that while models like Claude Mythos introduce new challenges, they also present opportunities for countries with strong technology talent pools — such as India — to lead in developing next-generation cybersecurity solutions.

The MGMM Outlook

The concerns raised by Nirmala Sitharaman highlight how rapidly advancing AI systems like Claude Mythos are beginning to blur the line between innovation and vulnerability. Developed by Anthropic, the model’s ability to detect and exploit deep-rooted software weaknesses signals a structural shift in cybersecurity risks, especially for sensitive sectors like banking. The possibility of such powerful tools falling into the wrong hands raises serious questions about the preparedness of existing digital infrastructure. India’s financial ecosystem, already heavily digitised, faces a new category of threats where attacks may become faster, more automated, and harder to detect using traditional security frameworks.

At the same time, the coordinated response involving institutions like the Reserve Bank of India and CERT-In reflects an urgent recognition that cybersecurity must evolve alongside AI capabilities. Strengthening real-time intelligence sharing, upgrading legacy systems, and building indigenous expertise in AI-driven defence are becoming essential priorities. Rather than treating such technologies purely as risks, there is a growing necessity to harness them strategically—turning potential threats into tools for resilience. This moment underscores a larger transition where technological leadership will depend not just on adoption, but on how effectively nations secure and govern these powerful systems.

(Sources: India Today, Indian Express, NDTV)