Lamentillo to HPAIR: AI mirrors human bias

CAMBRIDGE, Massachusetts — NightOwl AI founder Anna Mae Yu Lamentillo told delegates of the Harvard College Project for Asian and International Relations (HPAIR) that artificial intelligence programmes should not be treated as neutral or inherently fair, warning that systems built by people—and trained on human-generated data—can replicate the same inequities societies have long struggled to overcome.

In her address to HPAIR, Lamentillo argued that the “objectivity” often associated with AI is misleading, because the technology reflects the choices of its creators and the limits of the datasets that feed it. “AI can’t be equitable if whole communities are invisible in the data,” she said in a message shared alongside her HPAIR appearance.

HPAIR, a Harvard College student-run organization founded in 1991, convenes students and young professionals to discuss major issues facing the Asia-Pacific region through conferences held at Harvard and in the region, as well as a virtual conference. HPAIR’s Virtual Conference (VCONF) for Winter 2026 is scheduled for Feb. 13–15.

Lamentillo was recently featured by HPAIR as part of its VCONF 2026 speaker highlights, citing her work using AI to preserve endangered languages and fight digital exclusion.

Her remarks centered on how bias can enter AI systems through skewed representation—whether in the people building the tools, the priorities that shape product decisions, or the underlying data that may exclude minority communities, non-dominant languages, and marginalized groups. In a separate policy-focused speech published by Night Owl, Lamentillo pointed to “potential gender and social biases” linked to AI and urged policymakers to embed inclusion into regulatory processes.

Lamentillo’s advocacy draws from her work at NightOwl AI, a mission-driven initiative developing language technologies aimed at communities often left out of mainstream digital tools. In a published speech, she has described the widening gap created when advanced AI systems support only a fraction of the world’s languages, leaving many communities without meaningful access to education, services, and economic opportunities in an AI-driven world.

At HPAIR, she urged young leaders and future policymakers to treat AI governance as a question of human rights and representation—not just technical performance—calling for broader participation in dataset creation, stronger accountability standards, and deliberate investment in inclusion so that emerging AI products do not harden existing social and economic divides.

-Paraluman News