Safety · VentureBeat AI ·
The agent evaluation gap: Enterprise AI organizations have a reality-alignment problem, not a coverage problem — and most are shipping to production anyway
A VentureBeat Pulse survey of 157 enterprises found that 50% had deployed an AI agent or LLM feature that passed internal evaluations but later caused a customer-facing failure. Only 5% fully trust automated evaluations, while about two-thirds allow or are building toward production changes based on them without human