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 survey of 157 enterprises reveals a significant 'evaluation gap' where organizations grant AI agents increasing autonomy but lack trust in the evaluations meant to ensure reliability. Half have experienced agents passing internal tests only to fail in production, and only 5% fully trust automated evaluations. Despite this, two-thirds are moving toward fully automated deployment without human oversight, indicating autonomy is outpacing assurance.