The door is open: India's moment to shape global AI safety

There are three critical levers that Indian legislation can capitalise on immediately.

AI safety is having a moment.

The Indian government must seize this momentum to shape AI safety in a way that benefits us domestically and sets a global example for inclusive AI governance in emerging economies. (Representative Image/Unsplash)

Following last week’s resignation of Anthropic researcher Jacob Coxon and the public declaration to pace AI by the same company’s CEO, Dario Amodei, signed on to by tech CEOs like Sam Altman, Elon Musk, AI safety is getting an unprecedented moment in the spotlight. The Indian government must seize this momentum to shape AI safety in a way that benefits us domestically and sets a global example for inclusive AI governance in emerging economies.

There are three critical levers that Indian legislation can capitalise on immediately.

The first is the commitment to implementing the BRICS Global Governance agreement at the past weekend’s BRICS summit. The BRICS bloc has already achieved success in increasing IMF voting quotas and World Bank in favour of developing nations, in setting up a Contingency Reserve Arrangement to offer liquidity support during payment crises, and in local currency trade settlement. The message is also clear that it intends to act to amplify non-Western interests rather than directly work against the West.

This weekend’s New Delhi Declaration demonstrates the bloc’s commitment to firmly rejecting the weaponisation of AI that seems to be such a focus in the West, and look instead towards advancing inclusion and representation of its diverse populations. The Indian government must use this platform to learn from gains made in other BRICS nations (for example, Brazil’s General Law on Data Protection that prioritises guidance over sanctions in a phased approach), amplify our priorities in international fora, and lead as strong examples in fields we have an advantage in, such as digital public infrastructure and use of AI for financial inclusion.

Chinese premier Xi Jinping’s announcement of an open-source AI ecosystem at the BRICS forum is a strong opportunity for India to contribute and benefit from the shared development of lower-cost AI and AI capability building catered to BRICS country environments. It is also a platform to actively shape AI governance infrastructure by putting in place regulatory mechanisms around data and agents, boosting training on AI evaluations, and introducing an audit ecosystem around AI deployment. The buoyant techno-optimism around AI use in emerging markets needs a dose of legislative tempering to ensure its safe and inclusionary use.

The second lever is fintech. India’s fintech industry is the third largest and fastest-growing fintech market globally, with market capital around $120 billion and a fintech adoption rate of 87% (compared to a 64% global average). UPI transactions in January 2026 alone amounted to 21.7 billion transactions worth ₹28.33 lakh crore. With AI, and driven by rapid financial digitisation, the sheer magnitude of UPI, and a regulatory acceleration of digital lending and insurance distribution, that advantage is set to compound even more quickly, with estimates putting it at a 20.1% CAGR by 2036, the second-fastest globally by a close margin.

At the Global Fintech Fest 2026 last week, Finance Minister Nirmala Sitharaman called for AI guardrails and human oversight to meet increasingly sophisticated cyberattacks, the spread of misinformation, and risk of AI misalignment with human goals. A structure of good practices based on the lessons from the ongoing rollout of the EU’s landmark AI Act and situated in a regional context could help India position itself as a strong leader in countries with similar socioeconomic contexts. Sitharaman already proposed building a federated industry platform to help Indian startups navigate complex foreign licensing, data protection, cybersecurity, and tax regimes overseas. The sector should also look to build the architecture of regulation to safeguard its users and inspire trust on the international stage.

Existing mechanisms in place can propel this development. The Account Aggregator and Unified Lending Interface together remove the time and cost friction of financial flows to create individualised finance products for traditionally excluded microborrowers, often using AI. The RBI’s FREE-AI in turn enforces concrete compliance mechanisms that ensure unbiased and easily interpretable use of AI in such loan decisions. Scaling these mechanisms responsibly can lower costs, increase lending and build aggregated financial flows while also increasing financial inclusion. As more companies shift their financial flows through AI agents and agent interfaces, strengthening regulation around agentic AI payments could also help India to establish a first-mover advantage at a global scale.

Any advances in AI regulation will rest on the quality of the third lever, data governance. With one of the largest digital identity systems in the world, India has simultaneously an immense responsibility and opportunity in data governance. Advances in our digital government stack now link Aadhar details, mobile numbers, and even iris scans to all form of financial and logistical flows: among these, UPI payments, DigiYatra boarding cards, voter IDs, LPG connections. This has facilitated tracking and profiling highly disaggregated patterns in mobility, consumption, and even polling preferences in one of the largest populations in the world. In a world where data is currency and AI agents are increasingly able to jump ring fences to achieve assigned outcomes, it is essential to plug any weaknesses in data architecture in such a complex, interconnected system.

At the enterprise level, the Digital Personal Data Protection Act (DPDP) 2023, whose Rules were notified in November 2025 gives enterprises until May 2027 to comply in giving citizens control over the use of their digital data. Its enforcement will be critical to preventing its use in AI-powered cyberattacks and scams, misinformation campaigns, and the misuse of aggregated personal data outside of India’s borders. Care must be taken that similar protections are stringently applied to the interlocking gears of the Aadhar ecosystem to ensure that data neither leaks nor is misused or tampered with.

AI safety research from Model Evaluation & Threat Research (METR) reveals that the capabilities of frontier AI models double on average every 7 months, completely revamping the landscape in a way that calls for shorter, iterative review cycles every six months rather than the multi-year timelines typical of conventional legislation. The current call for slowing AI development by tech CEOs opens a brief window of prolongation for policy mechanisms to regroup and act in an agile, adaptive way.

India already has tenable examples of work being done on AI research. The IndiaAI Mission has already funded eight projects in its first funding round, from IIT Jodhpur's work on machine unlearning to Civic Data Labs' open source algorithm-auditing toolkit ParakhAI. It has also deployed 34,000 GPUs at fractional costs of that of AWS or Azure, reflecting India’s ability to operationalise AI safety research inexpensively and locally. Pioneering AI safety and governance efforts in India are represented by the inclusion of leading AI safety researcher Balaraman Ravindran at International Scientific Panel on AI, who is also behind drafting India's official AI Governance Guidelines and contributing to the RBI FREE-AI framework committee. Most recently, an AI Safety Institute established at the end of 2025 links academia, startups, industry and government replicates a structure existing in the West but for an Indian context.

India’s compute capacity remains modest by global standards. Public compute buildout, scaling towards 100,000 GPUs, is a fraction of what hyperscalers deploy, and the country continues to lose top AI research talent to AI labs in the US and UK. Strong governance frameworks built and tested now will serve as preparation to attract the compute and talent base that will eventually need to operate within them.

What India does in coming months will determine whether global AI governance and specific safety mechanisms account for our market structures and bring about financial inclusion at scale. Decisive and timely action now can serve as a bridge for trust in our markets and institutions while anchoring data protection firmly at the base.

The door for Indian policy to act is open. The question now is whether India will step through it.

Nanya Sudhir

The author is a former UN development economist building inclusive AI safety and governance

(AI Declaration: As a writer, I do not use and have not used AI to draft text. I sometimes use it to search for specific citations/quote origins, and to check if there are any logical gaps in my writing. I try to double check all facts offered by an LLM. All phrasing, logical flows, and potential errors are mine.)