Anthropic ban: Sarvam AI's Pratyush Kumar warns against reliance on foreign models

Kumar detailed that Sarvam has trained models at scale on roughly 3,400 Nvidia H100 GPUs and has brought India’s first Blackwell cluster online, targeting tens of megawatts of compute on Indian soil by 2027. After last week’s US government ban on Anthropic’s frontier models citing security risks, Sarvam AI cofounder Pratyush Kumar said Indian AI firms and users have been operating under a dangerous illusion.

The Argument

“Democratised AI sucks in all the data liquidity of the world which is locked in higher margin frontier offerings,” he wrote.

Further details indicate that “Bottomline, irrespective of hit jobs by foreign media or naysayers of various shades, India has the talent, the scale of the economy, the ambition amongst people for a better life, and a government that understands the imperative for technology innovation,” he concluded in his post.

“For AI users, it is clear that you should not confuse access with ownership, or adoption itself as advantage. And if the most significant tech differentiator you are leveraging has external control loops, then you have to accept you are vulnerable,”, Spokesperson

Why It Matters

After last week’s US government ban on Anthropic’s frontier models citing security risks, Sarvam AI cofounder Pratyush Kumar said Indian AI firms and users have been operating under a dangerous illusion. “For AI users, it is clear that you should not confuse access with ownership, or adoption itself as advantage. And if the most significant tech differentiator you are leveraging has external control loops, then you have to accept you are vulnerable,” he wrote in a post on X.

Analysis

As AI tooling proliferates, security teams face expanding attack surfaces tied to model inference and data pipelines.

The perspectives raised warrant consideration by both practitioners and decision-makers. While viewpoints may differ on the appropriate response, the underlying trends and data points deserve attention. Organizations are encouraged to evaluate how these insights apply to their specific context rather than adopting one-size-fits-all conclusions. Continued discourse and evidence-based debate remain essential for advancing the field.

Industry observers note that this type of development highlights the ongoing need for defense-in-depth strategies and proactive security posture management. Organizations that invest in regular security assessments and employee training tend to fare better when responding to emerging threats. The security community continues to share indicators and best practices to help defenders stay ahead.

Sources

1. https://economictimes.indiatimes.com/tech/artificial-intelligence/anthropic-ban-sarvam-ais-pratyush-kumar-warns-against-reliance-on-foreign-models/articleshow/131719431.cms