Tag
10 published articles
Safety is the condition of being protected from unacceptable risk, harm or injury in everyday life, workplaces, technology and public spaces. It involves identifying hazards, assessing likelihood and severity, and applying measures such as design standards, training, oversight and emergency planning. Safety also includes physical, digital, environmental and social dimensions, often balancing protection with usability, privacy and freedom.
Anthropic’s Model Hardware Standard is an early attempt to make AI-controlled laboratory and factory equipment more interoperable, but its real test will be whether technical guardrails can keep pace with physical risk.
Settlement discussions reported during Meta’s teen-addiction trial have culminated in a proposed multistate agreement that combines major payments with binding changes to Facebook and Instagram for young users.
Reports of students using generative AI to sexualise and impersonate educators show that deepfake abuse is becoming a workplace and safeguarding challenge for schools, not only a student-safety issue.
Fresh US Senate scrutiny has intensified a long-running debate over whether Roblox’s child-safety systems have kept pace with the risks created by a vast, highly monetised social gaming platform.
OpenAI’s latest security incident has turned a long-running debate over safety culture into a practical test of how a frontier AI company governs increasingly autonomous systems.
A New Mexico court has ordered Meta to contribute $567 million to a youth mental-health remediation fund, adding a major public-nuisance remedy to a separate $375 million consumer-protection penalty imposed in March.
A new investigation into Meta’s advertising systems shows how AI-enabled sexual exploitation can evade moderation even when platforms publicly describe child safety as a zero-tolerance priority.
World models and embodied AI could make machines more capable in the physical world, but dependable deployment will depend on advances in data, testing and safety as much as model scale.
The alleged shooting of licence-plate reader cameras in Tennessee has turned an online campaign about roadside safety into a warning about how surveillance disputes can create physical risks.
Google DeepMind’s Gemini Robotics 2 expands Gemini from digital reasoning into robot control, but its uneven task results show how far physical AI still has to travel.
