India's caste census debate exposes AI bias risk for fintech lenders

The readthrough for India's financial sector is direct. Banks and fintech lenders that rely on AI for credit scoring and underwriting face regulatory and reputational risk if their models reproduce caste bias indirectly. The RBI's 2021 warning already flagged the issue. A score that feels neutral can still embed discrimination, the Mint piece argued. An automated score spreads blame so evenly among the bank, vendor, model and employee that nobody can be held accountable, unlike a prejudiced clerk who can be accused.

Matrimonial platforms have rebranded old family filters, caste, gotra, complexion, horoscope, as compatibility science. Digitization did not invent prejudice, the piece said; it reduced its cost. A family that once vetted 20 candidates can now exclude 20,000 before breakfast, exclusion resembling consumer choice.

The labour market faces similar dynamics. Algorithms can assign a gig worker's route, set his pay, time his pauses and suspend him by SMS, with no one to appeal to. The hierarchy has become harder to spot but not easier to escape, the piece said.

The coming official caste record in Census 2027 adds another layer. Census records are confidential by law, but the exposure lies in what future governments or vendors might link, or what private systems may infer. If errors get wired into the digital infrastructure used by a sixth of humanity, an unjust decision could injure those who can least afford to appeal.

The article's author noted that those best placed to worry publicly about caste and AI are usually best placed to escape both. "The algorithm already knows your caste," the piece concluded. "The only open question is whether the Republic will catch up with what the market has been inferring for years."