Infosys CEO Sees ‘Mythos’ Risk As Opportunity

Infosys CEO Salil Parekh offered an unusually candid view on emerging risks in AI systems, referencing Anthropic’s Mythos vulnerabilities and positioning them as both a threat and a business opportunity.

Parekh says, based on current understanding, AI systems are revealing more vulnerabilities than previously thought.

While he did not go into technical specifics, the broader point was clear. As enterprises adopt AI at scale, new layers of risk are emerging around model behaviour, security, and reliability.

Parekh framed this as a potential growth area for Infosys.

He says they could help their clients by ensuring that vulnerabilities are identified and protected against quickly.

This aligns with Infosys’ broader AI positioning, which aims to move beyond implementation to governance, risk management, and the ongoing optimisation of AI systems.

The company’s partnerships with foundation model providers like OpenAI and Anthropic are expected to play a role here, especially as enterprises look for vendors who can manage both deployment and safeguards.

The comment also signals a shift in how IT services firms are thinking about AI. The conversation is no longer just about adoption, but about control, resilience, and trust.

Infosys ended FY26 with a milestone. Revenue crossed $20 billion for the first time ever, but the company guided for just 1.5%-3.5% growth in constant currency for FY27, underscoring a cautious demand environment even as AI-led deals pick up.

For the March quarter, Infosys reported revenue of ₹46,402 crore, up 13.4% year-on-year, but down 1.3% sequentially in constant currency terms. Full-year revenue stood at ₹1,78,650 crore, a 9.6% increase, translating to 3.1% growth in constant currency.