Who decides what AI tells you? Campbell Brown, once Meta's news chief, has thoughts

Former Meta news chief Campbell Brown argues that AI’s role as an information funnel poses systemic risks and that accuracy should be a priority for model makers. Her startup, Forum AI, benchmarks foundation models on “high‑stakes” topics (geopolitics, mental health, finance, hiring) by recruiting top domain experts and training AI judges to reach ~90% consensus with those experts. Brown says current models show political bias and factual failures (e.g., Gemini sourcing from CCP sites) and that existing compliance audits are inadequate. Forum AI is targeting enterprise customers — which care about liability in credit, lending and hiring — as its primary revenue path. For investors in Meta (META.OQ) and other AI‑focused firms, Brown’s critique highlights reputational and regulatory risks tied to misinformation, but also potential commercial demand for robust third‑party evaluation and compliance solutions.