New Delhi Frontier AI Impact Commitments: India’s Gift to Global AI Governance

February 19, 2026

At a pivotal moment for artificial intelligence, India has offered the world a balanced, pragmatic framework for responsible innovation.

On February 19, 2026, during the high-profile Leaders’ Plenary session at the India AI Impact Summit in Bharat Mandapam, a landmark voluntary framework was unveiled: the New Delhi Frontier AI Impact Commitments. Designed to guide developers, companies, governments, and researchers toward building responsible, transparent, and inclusive frontier AI systems, this framework represents India’s unique contribution to the global conversation on AI governance .

As powerful models approach—and in some domains exceed—human-level capabilities, the need for ethical guardrails has never been more urgent. The New Delhi Commitments rise to this challenge, offering principles that balance innovation with safeguarding humanity .

The Genesis: From MANAV Vision to Global Commitments

The New Delhi Frontier AI Impact Commitments didn’t emerge in a vacuum. They build directly upon Prime Minister Narendra Modi’s MANAV Vision (Moral and Ethical System, Accountable Governance, National Sovereignty, Accessible and Inclusive technology, Valid and Legitimate systems), unveiled just hours earlier at the same summit .

Where the MANAV Vision provides the philosophical foundation, the New Delhi Commitments offer a practical, actionable framework for developers and deployers of frontier AI systems. Together, they form a comprehensive approach to AI governance that prioritizes human welfare alongside technological progress.

As Union Minister Ashwini Vaishnaw eloquently described it: “This is a made-in-India contribution to the world—a balanced, pragmatic approach that promotes innovation while safeguarding humanity” .

The Six Pillars of the New Delhi Commitments

The voluntary framework outlines six core principles that participants are encouraged to adopt and implement:

1. Transparency & Explainability

The Commitment: Developers commit to disclosing model capabilities, limitations, training data sources (where feasible), and decision-making processes—particularly in high-stakes applications.

Why It Matters: As AI systems become more powerful, understanding their inner workings becomes essential for trust. When an AI model denies a loan, recommends a medical treatment, or influences a judicial decision, those affected have a right to understand why.

Implementation Examples:

Publishing model cards detailing capabilities and limitations
Disclosing the provenance and curation of training datasets
Providing explanations for AI decisions in high-stakes domains

2. Risk Assessment & Mitigation

The Commitment: Mandatory pre-deployment evaluations for frontier models, focusing on misuse risks, bias amplification, and potential societal harms.

Why It Matters: Frontier AI systems come with unprecedented capabilities—and unprecedented risks. From disinformation campaigns to autonomous weapons, from biosecurity threats to systemic bias, the potential for harm must be identified and mitigated before deployment.

Key Focus Areas:

Misuse risks: Disinformation, fraud, social manipulation
Security threats: Autonomous weapons, cyberattacks, biosecurity
Societal harms: Bias amplification, discrimination, erosion of social trust

3. Inclusive & Equitable Access

The Commitment: Emphasis on multilingual support, affordability, and solutions tailored for the Global South—ensuring AI benefits emerging economies, rural populations, and underrepresented groups.

Why It Matters: AI has the potential to either bridge or widen existing inequalities. The New Delhi Commitments explicitly prioritize inclusion, recognizing that the benefits of AI must reach beyond Silicon Valley boardrooms to farmers, small businesses, and students across the developing world.

Implementation Priorities:

Support for all 22 scheduled Indian languages and hundreds of global languages
Voice-first interfaces for populations with low literacy
Affordable pricing models for startups, researchers, and public institutions
Culturally relevant training data and applications

4. Ethical Guardrails

The Commitment: Alignment with human values, human rights, and democratic principles; commitments to red-teaming, safety testing, and ongoing monitoring post-deployment.

Why It Matters: As AI systems gain autonomy and influence, ensuring they remain aligned with human values becomes critical. The New Delhi framework emphasizes not just initial safety testing but continuous monitoring throughout a model’s lifecycle.

Core Elements:

Alignment with fundamental human rights
Respect for democratic processes and institutions
Comprehensive red-teaming before deployment
Ongoing monitoring for emergent behaviors post-deployment

5. Accountability & Reporting

The Commitment: Mechanisms for public reporting on safety practices, incident disclosure (when material), and collaboration with regulators and civil society.

Why It Matters: Accountability transforms abstract principles into concrete action. By committing to transparency about safety practices and incident disclosure, developers invite scrutiny and demonstrate their commitment to responsible development.

Key Mechanisms:

Regular public reports on safety practices and evaluations
Timely disclosure of material incidents or harms
Active collaboration with regulators and civil society organizations

6. Global Cooperation

The Commitment: Call for international alignment on frontier AI governance, knowledge-sharing, and joint standards—positioning India as a bridge between the Global North and South.

Why It Matters: AI knows no borders. A model developed in one country can impact citizens in another. Effective governance requires international cooperation, shared standards, and mutual learning. India’s unique position—as both a major AI developer and a representative of Global South perspectives—makes it an ideal bridge-builder.

Cooperation Priorities:

Harmonizing safety standards across jurisdictions
Sharing best practices for evaluation and testing
Ensuring Global South voices are heard in global governance forums

Early Signatories: A Who’s Who of Responsible AI

The New Delhi Frontier AI Impact Commitments carry significant symbolic and practical weight, with early signatories spanning Indian innovators, global tech giants, and government bodies:

Indian AI Leaders:

Sarvam AI —fresh off its landmark 105B-parameter model announcement
Krutrim —building sovereign AI infrastructure for India
Other Indian foundation model developers committed to responsible practices

Global Technology Companies:

Anthropic —known for its focus on constitutional AI and safety
OpenAI —represented at the summit, signaling global alignment
Google —committed to responsible AI development worldwide

Government and Institutional Support:

IndiaAI Mission —integrating commitments into national programs
Ministry of Electronics & IT —providing policy backing
NITI Aayog —supporting implementation and monitoring

This diverse coalition demonstrates that responsible AI governance is not a North-South or public-private divide—it’s a shared challenge requiring collective action .

Complementing India’s Emerging AI Architecture

The New Delhi Commitments don’t exist in isolation. They complement and reinforce India’s broader AI governance ecosystem:

IndiaAI Mission

The national mission provides compute subsidies, startup support, and institutional backing for homegrown AI innovation. By embedding the Commitments into its programs, it ensures responsible practices scale with technological capability.

Digital Personal Data Protection (DPDP) Act

India’s comprehensive data protection framework ensures that the data rights pillar of the MANAV Vision has legal teeth. The Commitments extend this privacy focus to AI-specific contexts.

Upcoming Digital India Act

The forthcoming legislation is expected to include provisions for algorithmic transparency, accountability, and grievance redressal—creating a regulatory environment where the Commitments can evolve into formal guidelines .

Why This Matters: India’s Unique Contribution

The New Delhi Frontier AI Impact Commitments represent something genuinely novel in global AI governance:

1. A Global South Perspective

Previous AI governance efforts have been largely led by Western nations and institutions. The New Delhi framework explicitly centers the concerns of emerging economies—multilingual access, affordability, rural inclusion, and sovereign data rights.

2. Voluntary but Meaningful

While the commitments are voluntary, they carry real weight. Signatories subject themselves to public scrutiny, civil society oversight, and peer accountability. In an industry where trust is increasingly valuable, such commitments differentiate responsible players.

3. Pragmatic and Balanced

Unlike frameworks that prioritize either innovation or safety exclusively, the New Delhi approach seeks balance. It recognizes that responsible development is sustainable development—and that guardrails, properly designed, can actually accelerate adoption and trust.

4. Evolving by Design

The Commitments are explicitly designed to evolve. As Minister Vaishnaw noted, they are intended to develop into more formal guidelines over time, learning from implementation experience and adapting to technological change .

The Road Ahead: From Commitment to Practice

Announcing principles is one thing; implementing them is another. The New Delhi framework includes pathways for moving from aspiration to action:

For Developers:

Conduct pre-deployment evaluations aligned with Commitments
Publish transparency reports detailing compliance
Establish internal accountability mechanisms
Engage with red-teaming and external audits

For Governments:

Integrate Commitments into procurement policies
Use them as benchmarks for public-sector AI deployment
Support multi-stakeholder forums for ongoing dialogue
Work toward international harmonization of standards

For Civil Society:

Monitor compliance and hold signatories accountable
Provide feedback on real-world impacts
Advocate for strengthening Commitments over time

For the Global Community:

Endorse the framework and encourage adoption
Share lessons learned from implementation
Collaborate on joint standards and mutual recognition

India’s Leadership Moment

The New Delhi Frontier AI Impact Commitments represent more than a policy announcement—they are a defining moment for India’s role in shaping the global AI future. By offering a framework that balances innovation with responsibility, inclusion with excellence, and sovereignty with cooperation, India has staked its claim as a thoughtful, principled voice in the most important technology conversation of our time.

As the India AI Impact Summit 2026 draws to a close, the message is clear: India isn’t just building AI—it’s helping build the ethical foundation upon which AI will serve all humanity.

The Commitments are now open for broader endorsement, inviting organizations worldwide to join this collective effort. For those building frontier AI systems, the question is no longer just “Can we build it?” but “Will we build it responsibly?” The New Delhi framework offers a path toward answering that question with integrity.