MIT's Michiel Bakker proposes using automated oversight to maintain democratic accountability over autonomous AI agents in public institutions

The talk explores how we can ensure that institutions increasingly run by agents remain accountable to democratic decisions and public commitments when the key evidence of decision-making resides within AI systems. It argues for automated oversight and will explore how it can be applied in frontier labs and beyond.

Find out more and register here: https://www.cooperativeai.com/seminars/closing-the-democratic-loop-automated-oversight-for-the-agi-era