US AI export restrictions show technology is ultimate weapon: Sridhar Vembu
New Delhi, June 13
Zoho co-founder and former CEO Sridhar Vembu on Saturday said recent US restrictions on access to advanced artificial intelligence models underscore the growing importance of technological self-reliance, asserting that technology has become central to national security and sovereignty.
Reacting to the US government's export controls that led to the suspension of Anthropic's advanced AI models -- Mythos and Fable -- for users outside the United States, Vembu said the development highlighted how strategic technologies are increasingly shaping geopolitical realities.
"This is big: all access to Mythos and Fable AI models disabled for everyone outside America," Vembu said in a post on social media platform X.
Calling technology the "ultimate weapon", he argued that national sovereignty and national security are now closely tied to technological capabilities.
"Globalisation is dead and Bharat must find her own way ahead," Vembu said, adding that the development should serve as a reminder for India to strengthen its domestic technology ecosystem.
He urged Indian organisations to adopt smaller AI models, including open-source models developed in India and China, saying they can be made effective with sufficient effort.
"Why pay money to people who don't even want to sell to you?" he remarked.
Vembu also called for deeper investments in research and development, noting that building cutting-edge frontier AI models requires enormous financial resources and access to advanced graphics processing units (GPUs), both of which remain concentrated in a handful of countries.
According to him, training the latest generation of AI models can require investments running into tens or even hundreds of billions of dollars, while access to the necessary computing infrastructure is increasingly subject to restrictions.
He said India should focus on alternative and cost-effective research approaches rather than attempting to match the spending levels of global AI leaders.
"Zoho has been pursuing alternative R&D approaches that are far less expensive," Vembu said, expressing confidence that sustained innovation and patience would eventually yield results.
The entrepreneur also said that while frontier AI development requires massive capital expenditure and access to restricted hardware, India can continue building capabilities through innovative approaches that are less resource-intensive.
Notably, AI company Anthropic announced that the US government had issued an export control directive ordering it to suspend access to its advanced "Fable 5" and "Mythos 5" models for foreign nationals, including its own employees, both inside and outside the United States.
The company said it was disabling access to the models to comply with the directive, while access to its other AI models remains unaffected.
Anthropic said it believes the decision stems from a misunderstanding and that it is working to restore access to the affected models as soon as possible.
— IANS
Reader Comments
Vembu makes a valid point but I'm cautiously optimistic. India has the talent pool, but we lack the massive capital investments in R&D that the US pours into frontier AI. Zoho's approach of cost-effective innovation is smart, but can it scale to compete globally? We need both government support and private sector collaboration.
This is a wake-up call for everyone in India's tech industry. We've been too comfortable with outsourcing and building on top of US platforms. Meanwhile China has invested billions in homegrown AI. We need our own GPUs, own models, own data centers. Bharat needs to wake up! 😤
Respectfully, I disagree with Vembu's 'globalisation is dead' take. Globalisation isn't dead, it's just being weaponised by the US. India should double down on strategic partnerships with Japan, EU and even China for AI chips and research. Going completely alone isn't feasible when frontier models need billions in investment.
The irony is that so many Indian engineers and researchers have built these US AI companies. Our human capital is world-class but we keep exporting it. Imagine if even half of the Indian diaspora in Silicon Valley came back or invested remotely in Indian AI startups. That would be a game changer for sure.
I completely agree with Vembu's point about smaller models. Not every Indian business needs a trillion-parameter model. Open-source alternatives like Llama, Mistral, and even Indian-developed models can do 90% of the job at 1% of the cost. Let the US spend billions on frontier models - we can innovate
We welcome thoughtful discussions from our readers. Please keep comments respectful and on-topic.