Prime Minister Narendra Modi unveiled the MANAV vision for artificial intelligence at the India AI Impact Summit 2026, introducing a comprehensive ethical framework that positions human welfare, accountability, and national sovereignty at the center of AI governance—a stark departure from Western models emphasizing market efficiency and Chinese approaches prioritizing state control.
The acronym MANAV, meaning “human” in Sanskrit, serves as both philosophical anchor and practical policy roadmap. The framework’s five pillars establish India’s claim to moral leadership in global AI governance while addressing legitimate concerns about algorithmic bias, data exploitation, and technological colonialism.
The Five Pillars
Moral and Ethical Systems: The first pillar demands that AI development be guided by ethical frameworks prioritizing human dignity and welfare over commercial profit or state surveillance. This is not mere rhetoric—the summit witnessed a Guinness World Record as 250,946 students pledged commitment to responsible AI development within 24 hours, demonstrating grassroots engagement with ethical considerations.
“Ethics cannot be an afterthought or a public relations exercise,” Modi emphasized in his keynote address. “MANAV demands that moral considerations are embedded in AI architecture from the first line of code.”
This pillar mandates ethical review boards for high-risk AI applications, analogous to institutional review boards in medical research. Government AI deployments in healthcare, criminal justice, and social services must demonstrate ethical clearance before implementation.
Accountable Governance: The second pillar establishes transparent oversight mechanisms through the ₹10,300 crore IndiaAI Mission and associated governance guidelines. Unlike opaque corporate AI development or state-controlled systems, MANAV requires public accountability for AI systems affecting citizens.
Union Minister Ashwini Vaishnaw elaborated: “Accountability means citizens can question how decisions affecting them are made. If an AI system denies someone a loan or flags them for investigation, they have the right to understand why and challenge that decision.”
The framework mandates algorithmic impact assessments for government AI systems, requiring documentation of training data sources, performance metrics across demographic groups, and failure mode analysis. These assessments must be publicly accessible, subject to independent audit.
National Sovereignty: Perhaps the most geopolitically significant pillar, national sovereignty asserts India’s right to control data generated by Indian citizens and govern AI systems operating on Indian territory. The principle “Whose data, their rights” directly challenges data extractivism practiced by multinational technology corporations.
“For too long, Indian data has been mined to train foreign AI systems that then sell services back to us at premium prices,” Modi stated. “MANAV declares that Indian data belongs to Indians and must primarily benefit Indians.”
This pillar has provoked anxiety among Western technology companies concerned about data localization requirements potentially fragmenting the internet. However, Indian officials argue sovereignty is non-negotiable given historical patterns of colonial resource extraction now manifesting digitally.
The India Semiconductor Mission, discussed in Report 4, provides the hardware foundation for technological sovereignty. The homegrown AI models detailed in Report 3 represent software sovereignty. MANAV provides the governance framework integrating these elements into coherent national strategy.
Accessible and Inclusive: The fourth pillar commits India to ensuring AI functions as a “multiplier, not a monopoly”—technology amplifying human capability rather than replacing human agency or concentrating power in elite hands.
Practical manifestations include the democratized GPU access discussed in Report 2, making computational infrastructure available at ₹65 per hour rather than market rates exceeding ₹500 per hour. The pillar also mandates multilingual AI systems, ensuring India’s linguistic diversity is supported rather than erased by English-centric technology.
“Inclusion is not charity—it is prerequisite for legitimate AI governance,” Modi argued. “AI that serves only urban, English-speaking elites while marginalizing rural, vernacular-speaking majority is not progress—it is technological apartheid.”
The accessibility mandate extends to persons with disabilities. The summit showcased 18 AI innovations improving autonomy for persons with limitations, developed through collaboration between ALIMCO and IIIT Bangalore. These include AI-powered navigation aids, real-time sign language translation, and voice-controlled home automation optimized for Indian contexts.
Valid and Legitimate: The final pillar requires AI systems to be lawful, verifiable, and free from fabricated content. This addresses escalating concerns about deepfakes, disinformation, and synthetic media undermining truth itself.
Modi’s call for global standards on content authenticity—likening them to nutrition labels on food—reflects growing consensus that trust must be built into AI systems architecturally, not retrofitted after deployment.
“A photograph used to be evidence. A video used to be proof. AI has destroyed that certainty,” Modi observed. “We must rebuild trust through authentication mechanisms embedded in every AI-generated image, video, and audio.”
India’s 2026 rules on deepfakes mandate three-hour takedown requirements for platforms hosting synthetic media lacking authenticity watermarks. While some civil liberties advocates worry about censorship implications, the government maintains that verifiable provenance is essential for democratic discourse.
Guinness Record: Symbolic or Substantive?
The 250,946 student pledges for responsible AI represent the largest collective commitment to technology ethics in history, earning Guinness World Records certification. Critics dismiss this as performative symbolism divorced from meaningful impact.
“Pledges are easy. Implementation is hard,” cautioned Dr. Sunil Verma, technology policy researcher. “The question is whether these students, once employed by companies prioritizing profit over ethics, will maintain those commitments when pressured to compromise.”
However, proponents argue the exercise serves pedagogical value, forcing students to articulate ethical principles and consider AI’s societal implications before entering industry. The pledges form part of curriculum in over 1,200 engineering colleges nationwide, ensuring sustained engagement with MANAV principles.
Contrasting Governance Models
MANAV positions itself as alternative to dominant AI governance paradigms. The Western model, exemplified by Silicon Valley, emphasizes market-driven innovation with minimal regulation, trusting corporate self-governance and consumer choice to produce optimal outcomes.
This approach has generated remarkable innovation but also profound failures: algorithmic bias in hiring systems, surveillance capitalism monetizing personal data, and concentration of AI capability in handful of corporations controlling access to foundational technologies.
China’s model emphasizes state control and techno-nationalism, with AI development explicitly subordinated to Communist Party priorities. This enables rapid deployment but raises concerns about surveillance, social credit systems, and suppression of dissent through algorithmic governance.
MANAV attempts a third path: state-guided development oriented toward social welfare rather than commercial profit or state control, with democratic accountability mechanisms and emphasis on human agency.
“We reject the false choice between unregulated corporate AI and authoritarian state AI,” explained S. Krishnan, Secretary of MeitY. “MANAV demonstrates that democratic governance can guide technological development toward genuine human flourishing.”
Implementation Challenges
MANAV’s lofty principles face formidable implementation obstacles. India’s regulatory infrastructure remains fragmented across multiple agencies with overlapping jurisdictions and frequent turf battles. The Information Technology Act of 2000 lacks provisions for AI-specific challenges like algorithmic accountability or synthetic media.
Legal expert Dr. Pawan Duggal notes that India’s legal system struggles with technologies moving orders of magnitude faster than legislative processes. “We are applying 2000-era laws to 2026 technologies,” Duggal observed. “MANAV articulates beautiful principles, but without legislative teeth, they remain aspirational.”
Additionally, enforcement capacity remains questionable. India’s technology regulators are understaffed and lack technical expertise to audit complex AI systems. The Centre for Development of Advanced Computing (C-DAC), designated as technical authority for AI auditing, employs fewer than 50 AI specialists—insufficient to oversee thousands of AI deployments nationwide.
Resource constraints also challenge accessibility commitments. While subsidized GPU access serves 600 startups showcased at the summit, India has over 50,000 registered technology startups. Scaling subsidized infrastructure to serve even 10% of that population requires investments potentially exceeding current IndiaAI Mission allocations.
Data Sovereignty
MANAV’s national sovereignty pillar has triggered fierce debate about economic implications. Multinational technology corporations argue data localization requirements increase operational costs, reduce efficiency, and fragment the global digital economy.
“Data flows are the lifeblood of modern AI,” argued a senior executive from a major U.S. technology company speaking anonymously due to diplomatic sensitivities. “Forcing data localization is like forcing international airlines to use different aircraft for each country. It’s inefficient and ultimately self-defeating.”
However, Indian officials counter that data sovereignty protects strategic interests and ensures value capture. “When Indian users generate data that trains American AI models sold back to India, that’s digital colonialism,” argued Krishnan. “Data sovereignty ensures value remains in India.”
Economic modeling suggests competing outcomes. Pessimistic scenarios predict data localization could reduce India’s digital economy growth by 0.7-1.2 percentage points annually through 2030 as multinationals reduce investment. Optimistic scenarios project increased investment in domestic data infrastructure and AI development, potentially adding 1.5-2.0 percentage points to digital economy growth.
The actual outcome likely depends on implementation details and whether India’s domestic AI industry develops competitiveness justifying reduced access to foreign platforms.
The Global South Leadership Bid
MANAV positions India as champion of Global South interests in AI governance. The framework’s emphasis on equity, accessibility, and development rather than existential risks resonates with nations prioritizing immediate socioeconomic challenges over speculative dangers.
“The West wants to discuss whether AI might someday become sentient and destroy humanity,” observed Brazil’s President Lula during his summit address. “The Global South wants to discuss how AI can improve agricultural yields today. MANAV focuses on what matters to us.”
This philosophical alignment has strategic implications. If Global South nations adopt MANAV principles, India shapes global governance norms through standard-setting rather than coercion—exactly how the United States and European Union have historically exercised technological soft power.
However, India’s leadership faces challenges from China’s aggressive AI diplomacy, particularly through Digital Silk Road initiatives providing AI infrastructure to developing nations. While India offers principles, China offers hardware. That tangible advantage proves compelling for resource-constrained nations.
Content Authenticity
Modi’s proposal for universal content authenticity standards—effectively “nutrition labels” for AI-generated content—raises complex technical and political questions. Multiple competing authentication schemes exist: digital signatures, blockchain-based provenance, cryptographic watermarking embedded imperceptibly in media.
No consensus exists on which approach should become global standard. Additionally, enforcement remains problematic. Any authentication scheme can be circumvented by bad actors willing to strip metadata or recreate content without watermarks. The technology resembles digital rights management (DRM) for movies and music—theoretically sound but practically defeated by determined adversaries.
Moreover, authentication requirements could enable censorship. If all content must carry authenticated provenance, anonymity becomes impossible. This threatens whistleblowers, political dissidents, and vulnerable populations requiring anonymous communication for safety.
“Content authenticity is crucial for trust, but implementation must not sacrifice anonymity where legitimately needed,” cautioned Apar Gupta, executive director of the Internet Freedom Foundation. “MANAV must balance competing values rather than maximizing single dimension.”
Open AI Advocacy
MANAV’s emphasis on open AI—shared code and wider participation enabling safer, more inclusive systems—reflects India’s historical commitment to knowledge as public good rather than private property.
“The Sanskrit tradition views knowledge as divine gift meant to benefit all humanity, not proprietary asset enriching individuals,” Modi observed, drawing on civilizational heritage. “Open AI aligns with this tradition while providing practical advantages in security, innovation, and accessibility.”
Research supports open source advantages in security through transparency enabling collective debugging. Conversely, closed proprietary systems hide vulnerabilities until catastrophically exploited. Open AI also accelerates innovation by allowing researchers globally to build upon shared foundations rather than redundantly recreating basic infrastructure.
However, the open versus closed debate remains contentious even within India. As Report 3 detailed, Sarvam AI chose proprietary approach for commercial sustainability while BharatGen embraced full openness. MANAV acknowledges both models while expressing philosophical preference for openness.
From Declaration to Reality
MANAV’s ultimate significance depends on translation from principles to practice. This requires legislative action, regulatory capacity building, international diplomacy, and sustained political will across electoral cycles.
The framework has achieved immediate diplomatic success, with 88 nations endorsing New Delhi Declaration principles incorporating MANAV elements. Whether this translates to concrete policy adoption remains uncertain.
Domestically, MANAV faces tests as government deploys AI in sensitive domains like criminal justice, social services, and surveillance. When efficiency gains conflict with ethical constraints, whether MANAV principles hold or yield to expedience will determine if the framework represents genuine commitment or rhetorical cover.
“Every governance framework sounds beautiful in pronouncement,” observed Dr. Verma. “The measure comes when hard choices arise between principle and convenience. Will India sacrifice efficiency for ethics? Will it accept economic cost for sovereignty? Those questions determine if MANAV is landmark or mirage.”
For now, India has staked claim to moral leadership in global AI governance. The summit audience of global leaders, CEOs, and civil society representatives signaled receptivity to India’s vision. Whether MANAV reshapes global norms or remains distinctly Indian experiment will emerge over the decade ahead as artificial intelligence pervades every dimension of human existence.
– Raja Aditya
