Who Decides What AI Says? Lessons from Campbell Brown and Forum AI

Colleagues, I’d like to share an observation from the AI field.

Campbell Brown, former head of news at Meta, founded Forum AI to evaluate how large models handle "high‑risk" topics — geopolitics, health, finance and hiring.

Forum AI convenes leading experts, builds benchmarks and trains "AI judges"; its target is roughly 90% agreement with human evaluators.

Brown highlights systemic failures: political bias, context loss and reliance on inappropriate sources.

Why it matters: businesses’ decisions and user trust hinge on AI response quality.

Which do you find more effective for assessing AI — external audits or expert-driven benchmarks?

#AI #AIEthics #ResponsibleAI #Tech