AI for Good

2026-02-10 · Source: AI Now Institute · Field: Technology & Digital — Artificial Intelligence & Machine Learning, AI Ethics & Governance · Depth: Intermediate, medium

Summary

Abeba Birhane, founder of the AI Accountability Lab and assistant professor at Trinity College Dublin, critiques the "AI for good" framework, arguing it technocratically addresses complex sociopolitical issues and obscures harmful industry practices. This framing, often adopted by large corporations like Microsoft and Google and international organizations, is seen as a "shiny veneer" over bad data, exploitative practices, and weak evidence. Birhane contends that AI tools cannot solve fundamental problems like hunger or gender violence, which require political will and systemic restructuring. She highlights how current AI systems often encode and exacerbate inequalities, while major tech companies promoting "AI for good" simultaneously contribute to environmental destruction and social regression. Instead of supporting Big Tech's initiatives, Birhane advocates for investing in smaller, community-based efforts that genuinely serve local needs without making grandiose claims, and urges governments to demand sound empirical evidence for AI's claimed social benefits.

Key takeaway

For policymakers and organizations considering "AI for good" initiatives, you should critically evaluate claims by demanding empirical evidence of social benefit, rather than relying on potential or promises. Prioritize supporting smaller, community-based technology efforts that demonstrate tangible positive impact over large-scale, corporate-driven projects that may obscure underlying harms or perpetuate existing inequalities. Your decisions should be based on what genuinely benefits people, especially marginalized communities, and the environment.

Key insights

The "AI for good" narrative often masks technosolutionism, weak evidence, and harmful practices by large tech corporations.

Principles

• Sociopolitical issues require political will, not just technological solutions.
• AI systems can encode and exacerbate existing inequalities.
• Empirical evidence should validate AI's claimed social benefits.

In practice

• Support small, community-driven tech initiatives.
• Demand robust evidence for AI's positive impact claims.

Topics

• AI for Good Critique
• AI Accountability
• Technosolutionism
• Community-Based AI
• AI Governance Policy

Best for: AI Ethicist, Policy Maker, AI Researcher

Related on AIssential

• Four Experts on the Questions We Should Be Asking About AI Right Now — AI's evolving impact necessitates proactive, cross-sectoral collaboration to address long-term societal changes, equitable benefits, knowledge integrity, and information overload.
• AI economics for the common good — AI's economic gains risk concentrating wealth, necessitating interventions to broaden benefits for global public good.
• Beyond Impact Lingo: Questioning, Concretizing, Building — AI "impact lingo" often masks environmental exploitation, exclusion, and imperialism, necessitating critical reframing.
• Bernie Sanders Saw This Coming — Sanders argues AI's transformative power necessitates public ownership and regulation to prevent oligarchic control and ensure societal benefit.
• OpenAI Proposes a New Deal — AI's rapid advancement and negative public sentiment necessitate clear, actionable policy and communication strategies.
• Open Source — Open-source AI often misrepresents its capabilities, failing to address core power concentrations in the AI industry.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI Now Institute.