#manipulation

Wired analyzes Eka's robotic claw technology, suggesting it could represent a breakthrough moment for physical manipulation akin to ChatGPT's impact on text AI.

Google DeepMind says it has built a harmful manipulation evaluation toolkit from nine studies spanning more than 10,000 participants. The work argues that manipulation risk is domain-specific, with finance and health producing very different outcomes.

Google DeepMind said on March 26, 2026 that it is releasing research on how conversational AI might exploit emotions or manipulate people into harmful choices. The company says it built the first empirically validated toolkit to measure harmful AI manipulation, based on nine studies with more than 10,000 participants across the UK, the US, and India.