As Anthropic restricts access, Indian firms scramble to test a model capable of exposing hidden vulnerabilities
A High-Stakes Scramble for Access
India’s leading fintech firms are urgently seeking early access to Anthropic’s Mythos, an AI model feared for its potential to reshape cybersecurity risks.
Companies like One97 (Paytm), Razorpay, and Pine Labs want to test their systems before adversaries do.
Is this proactive defense—or a signal that the threat has already outpaced safeguards?
- Mythos can reportedly detect vulnerabilities undetected for years
- Access currently restricted to a small, vetted group of companies
Why Mythos Has Triggered Global Alarm
Anthropic has labeled Mythos too powerful for broad release, rolling it out selectively under Project Glasswing.
The model’s capabilities have rattled regulators and policymakers worldwide.
How dangerous is a system that can uncover systemic weaknesses across financial infrastructure?
- US Treasury Secretary Scott Bessent called it a “step function change”
- ECB President Christine Lagarde warned of misuse risks
- Could expose banking and payment system vulnerabilities at scale
Indian Fintechs: A Race Against Time
Executives describe the situation as urgent and existential.
Razorpay CEO Harshil Mathur said firms are working overtime, treating Mythos access as critical to strengthening defenses.
If attackers gain similar tools first, what happens to digital trust?
- Firms want to stress-test their own platforms
- Discussions around Mythos dominating startup circles
Anthropic’s Tight Control: Who Gets In—and Why
Access decisions are highly scrutinized.
Anthropic is reportedly questioning applicants on intended use cases and potential benefits before granting entry.
Does restricting access reduce risk—or concentrate power among a few players?
- Initial access includes AWS, Apple, JPMorgan
- Expansion planned with strict usage contracts
- Likely limited to non-commercial security testing
The Bigger Fear: Systemic Financial Risk
The anxiety extends beyond individual companies to entire economies.
Vijay Shekhar Sharma warned that AI-driven attacks could infiltrate systems without traditional warfare.
Could AI turn cyberattacks into the new form of geopolitical conflict?
- Risks include mass account breaches
- Potential disruption of global payment networks
- Concerns over financial system stability
India’s Strategic Position in the AI Ecosystem
India plays a critical role, with millions of engineers supporting global financial systems.
It is also the second-largest market for Anthropic’s Claude, largely for coding and system modernization.
Does this make India more resilient—or more exposed?
- Deep integration with global financial infrastructure
- Heavy reliance on AI-assisted development workflows
Regulation and the Road Ahead
Experts expect tighter regulatory oversight as risks escalate.
Pine Labs CEO Amrish Rau emphasized that security must evolve beyond a compliance checkbox.
Will regulators move fast enough to match AI’s pace?
- Anticipated stricter cybersecurity norms
- Shift toward continuous, proactive defense models
TL;DR
Indian fintech firms are pushing for early access to Anthropic’s Mythos, an AI model capable of uncovering deep cybersecurity flaws. With global regulators alarmed, companies see access as critical to defense. The situation highlights rising fears that advanced AI could destabilize financial systems if misused.
AI summary
- Indian fintechs seek Mythos access urgently
- Model can detect long-hidden vulnerabilities
- Anthropic restricting rollout via Project Glasswing
- Global regulators warn of systemic risks
- Security shifting from compliance to proactive defense
