India’s new rules for AI governance are built on principles, not hard rules on what is and isn’t allowed. That implies less trouble with following the rules and more room to ship goods for startups. For big tech, it means higher demands for safety, accountability, transparency, and local responsibility.
What the 7 Sutras mean
These directives stretch across seven sutras, whose deeper meaning includes: Trust as the Foremost Principle, Human Beings Before Markets, Innovation is a Greater Good than Restriction, Justice and Equity, Liable by Nature, Design Must be Comprehensible, and the question of Trust and Responsibility and Sustainability. Collectively, they provide the framework on how AI must be designed and implemented in India — human-centered, inclusive, explainable and safe. The government has stated that the approach is intended to help “AI for All” and inclusive growth instead of curbing innovation with rigorous pre-approval rules.
Meaning for startups
The most significant implication for startups is that India is sending a message of support (rather than overregulation). Sutra: “Innovation over Restraint” — useful AI products should not be hindered by compliance layers that are not warranted from a risk perspective. Meanwhile, startup will have to bake some fundamental governance into early: responsibility, user clarifications, appeal procedures, and safety measure. A standalone AI law would not lead to more scrutiny than a sector regulator would of a startup using AI in finance, health, hiring, or public-facing automation.
Meaning for big tech
They raise the bar when it comes to trust and defensibility for large platforms and model providers. Accountability means organizations need to demonstrate who is responsible for what across the AI value chain — not just that the model is effective. “Understandability by Design” nudges companies away for less mystifying disclosures and more explainable outputs whenever possible. It matters even more for large players, since they are more likely to operate high-impact systems, large scale compute, and critical infrastructure integrationsfor “Safety, resilience and sustainability.”
What companies should do
In practical terms, this translates into you should document more, explain more, and monitor more. Companies ought to map AI use cases by risk, inject human oversight for sensitive decisions, keep incident logs, and develop transparency reports or internal audit trails. They should also review data governance, bias testing, security controls, and fallback mechanisms, as the framework categorises these as responsible deployment, not optional extras. Additional recommendations like the AI Governance Group, the Technology & Policy Expert Committee and the AI Safety Institute are also proposed by India and this means the companies will experience a more synchronized policy ecosystem down the line.