Pentagon CTO says U.S. government shouldn't take stakes in tech giants, questions adding AI rules20m ago Economists Concerned the AI Bubble Is About to Blow24m ago Scoop: U.S. open to discuss AI "shared risks" with China, Bessent says33m ago Trump admin says private sector can solve AI threats as critics balk36m ago Rune Raises $40M to Turn Stranded Solar Power Into AI Compute41m ago MSK and OpenEvidence Announce Two-Way Oncology AI Integration42m ago Google will now let any AI agent run your smart home1h ago Mark Zuckerberg Takes Aim at Anthropic in Debate Over A.I. Slowdown48m ago Google DeepMind launches the DeepMind Institute to debate AGI51m ago Sen. Blumenthal urges AI oversight: 'We're on the verge of losing control'52m ago AI environmental concerns build as lawmakers grapple with tech panic53m ago House speaker calls early recess before midterms amid AI regulation frenzy59m ago Claude comes for Gemini with its own take on Docs and Slides2h ago At UN, Nations Wrestle With the Rise of Autonomous Weapons1h ago

Pentagon CTO says U.S. government shouldn't take stakes in tech giants, questions adding AI rules20m ago Economists Concerned the AI Bubble Is About to Blow24m ago Scoop: U.S. open to discuss AI "shared risks" with China, Bessent says33m ago Trump admin says private sector can solve AI threats as critics balk36m ago Rune Raises $40M to Turn Stranded Solar Power Into AI Compute41m ago MSK and OpenEvidence Announce Two-Way Oncology AI Integration42m ago Google will now let any AI agent run your smart home1h ago Mark Zuckerberg Takes Aim at Anthropic in Debate Over A.I. Slowdown48m ago Google DeepMind launches the DeepMind Institute to debate AGI51m ago Sen. Blumenthal urges AI oversight: 'We're on the verge of losing control'52m ago AI environmental concerns build as lawmakers grapple with tech panic53m ago House speaker calls early recess before midterms amid AI regulation frenzy59m ago Claude comes for Gemini with its own take on Docs and Slides2h ago At UN, Nations Wrestle With the Rise of Autonomous Weapons1h ago

Robotics AI Research6 min reading time

Rethinking Robot Safety in the Age of AI

IEEE SpectrumPublished 1 hour ago

Recent research reveals that AI-powered robots can be manipulated through hidden triggers in their sensory inputs, causing unsafe actions despite normal operation otherwise. These vulnerabilities challenge traditional robot safety assessments and highlight risks in AI perception and decision systems.

#adversarial-attacks #badnets #badvla #goba #neurips #robot-safety

More in Robotics

Agility Launches Digit 5: No More Safety Cages, $300 Million In Orders

Covered by 4 sources1d ago

Robotics4 min read

Robots are waiting for a ChatGPT moment: Nvidia’s Les Karpas explains why at TechCrunch Disrupt 2026

TechCrunch3h ago

Is There Anything Google’s Fruit Fly Brain Can’t Do?

New York Times5h ago