Key Takeaways
- Zscaler founder Jay Chaudhry criticizes US restrictions on advanced AI systems as a knee-jerk reaction
- Chaudhry says concerns about restricted access to Anthropic AI models are overblown
- Zscaler was among 40 companies granted early access to Anthropic's Mythos model under Project Glasswing
Jay Chaudhry, the Indian-origin founder and CEO of cybersecurity giant Zscaler, has labeled US restrictions on advanced AI systems a "knee-jerk reaction." Speaking to the Economic Times, Chaudhry dismissed fears over limited access to frontier AI models from companies like Anthropic, calling the concerns "overblown."
The comments land at a tense moment. Washington has steadily tightened controls on AI exports, citing national security risks tied to China's rapid advances in artificial intelligence. Tech executives have largely stayed quiet, wary of antagonizing regulators. Chaudhry is breaking from that script.
What did Chaudhry actually say?
In an interview with ET's Technology Editor Surabhi Agarwal, Chaudhry took direct aim at current AI policy. He argued that limiting enterprise access to advanced models will hurt American competitiveness without meaningfully slowing adversaries. The restrictions, in his view, are reactive rather than strategic.
“The US restrictions around advanced AI systems are a knee-jerk reaction.”
— Jay Chaudhry, Founder and CEO, Zscaler
Chaudhry's Zscaler sits at the intersection of two policy hot spots: AI and cybersecurity. The company provides zero-trust cloud security to over 8,000 enterprises globally and generates roughly $2.3 billion in annual revenue. That scale gives Chaudhry unusual visibility into how AI restrictions play out in practice across multinational corporations.
Why Anthropic matters here
Zscaler was one of just 40 companies selected for early access to Anthropic's advanced AI model, Mythos, through a program called Project Glasswing. That exclusive club suggests Zscaler has meaningful AI integration plans. It also means Chaudhry has firsthand experience with the hoops companies must jump through to access frontier models under current rules.
Anthropic, backed by Google and valued at over $18 billion, has faced its own regulatory pressure. Recent US orders have required the company to disable certain AI capabilities in specific regions. Enterprise customers caught in those restrictions are watching their AI roadmaps get rewritten by policy rather than product strategy.
The business case against AI export controls
Chaudhry's criticism reflects a growing frustration among enterprise tech leaders. The argument runs like this: blanket restrictions slow down legitimate business use cases without stopping determined state actors. China is building its own frontier models. Export controls primarily hamstring American companies and their global customers.
There's a counterargument, of course. National security officials point to dual-use risks. Advanced AI models trained on proprietary data could theoretically be repurposed for surveillance, disinformation, or weapons development. The question is whether current controls actually address those risks or just create compliance headaches.
Chaudhry clearly lands on the side of "compliance headache." But he's not alone. Several enterprise AI vendors have privately voiced similar concerns, though few have been willing to say so on the record.
What this means for Indian enterprises
India sits in an awkward spot in the global AI access debate. The country is a major market for US enterprise software and a significant source of tech talent. Yet Indian companies face the same access restrictions as firms in less friendly jurisdictions.
Chaudhry, as an Indian-origin founder with deep ties to both countries, occupies a unique vantage point. His public criticism may signal that more tech leaders will begin pushing back on policies they see as counterproductive.
For Indian CTOs evaluating AI vendors, the uncertainty is real. Will the model you build on today be restricted tomorrow? Chaudhry's comments suggest enterprise customers are already asking that question.
Logicity's Take
Chaudhry is saying out loud what many enterprise tech executives have been muttering privately. The question is whether his comments represent a turning point in the industry's public posture on AI regulation, or just an outlier willing to take the political risk. Either way, his access to Anthropic's Mythos through Project Glasswing suggests he has concrete experience with how restrictions affect real deployments. That lends credibility to his critique, even if it won't change policy tomorrow.
Frequently Asked Questions
What did Jay Chaudhry say about US AI restrictions?
Chaudhry called US restrictions on advanced AI systems a "knee-jerk reaction" and said concerns over limited access to Anthropic's models are "overblown." He made the comments in an interview with the Economic Times.
What is Project Glasswing?
Project Glasswing is a program through which Anthropic granted early access to its advanced AI model, Mythos, to approximately 40 select companies. Zscaler was among those chosen for access.
Why are US AI export controls controversial?
Critics argue the restrictions hamper legitimate enterprise use cases without stopping state adversaries who are developing their own AI capabilities. Supporters say the controls address national security risks from dual-use AI technology.
How does this affect Indian companies?
Indian enterprises face the same access restrictions as firms in other countries, creating uncertainty around which AI models will remain available for long-term deployments.
Need Help Implementing This?
Navigating AI vendor selection amid shifting export controls requires careful planning. Contact Logicity's enterprise advisory team for guidance on building AI strategies that account for regulatory uncertainty.
Source: Tech-Economic Times
Huma Shazia
Senior AI & Tech Writer
Produced with AI assistance and reviewed by the Logicity editorial team. Learn more in our Editorial Policy.
