In Good Faith: To build trust, AI needs to be anchored by human agency

Last month, India hosted the global AI summit. This week, there are sleepless nights in Delhi over the gathering war in West Asia and the plight of 10 million Indians there. One of the key themes that emerged at the AI summit was how the human element and human agency remain fundamental to trust and accountability in systems. That trust has resurfaced in the current discourse around the war as Anthropic’s Dario Amodei pushes back against the US government on using AI for surveillance and in the battlefield. As the world grapples with these threats, there’s a different path — one rooted in both the quiet progress of domestic reform and the unyielding moral compass of the individual. These two forces could ensure an ethical core in our most powerful technology.

Three recent examples illustrate this idea that trust is forged through the alignment of law with justice. One significant milestone in this pursuit is the Repealing and Amending Bill, 2025. By stripping away outdated and discriminatory colonial-era laws , the state signalled a commitment to modernising justice and governance and reshaping the compact between citizen and state. An example is the amendment to the Indian Succession Act, 1925. For decades, Hindus, Buddhists, Sikhs, Jains, and Parsis in the former presidencies of Bombay, Madras, and Calcutta were burdened by the discriminatory requirement to obtain court probate for their wills. The omission of Section 213 has removed this geographical and religious disparity, bringing tangible relief to millions. This is the “quiet progress” of reform — the dismantling of archaic barriers to ensure the law reflects the spirit of equality and secures inheritance for the future.

Advertisement

While policy provides the framework, individual acts of courage offer the soul. Consider the reported story of Haji Akhtar, a Faridabad scrap dealer who discovered Rs 15 lakh worth of gold misplaced in a sack of junk. Akhtar’s decision to return the jewels — without the gaze of a CCTV camera over him or the pressure of the police — was a triumph of an ethics rooted in daily lived experience rather than in a value imposed or programmed. This, when a Chinese robot was being passed off as an indigenous one at a summit stall.

Bridging the gap between individual acts of honesty, the looming power of AI, and policy lies at the core of the MANAV framework unveiled at the AI summit. MANAV is a reminder to act with accountability and morality in every digital interaction. It reaffirms that inclusivity ensures justice is transparent and caring, and underlines the sovereignty of the self, ensuring humans remain “wise beings” rather than mere appendages to robots.

In this framework, for AI to be successful, it must adjust to a world of billions of “Manavs” (humans) who, whether literate or illiterate, refuse to fail in their hosting of honesty.

Advertisement

Nilay is the author of Being Good, Aaiye, Insaan Banen and Ethikos. He teaches and trains courses on ethics, values and behaviour