Zoox, under Amazon’s umbrella, has initiated a voluntary software recall for its robotaxis after a collision in Las Vegas, highlighting the ongoing challenges in autonomous vehicle safety.
Tag: AI Safety
Google’s internal benchmarks reveal that the Gemini 2.5 Flash AI model performs worse on safety tests compared to its predecessor, with notable regressions in text-to-text and image-to-text safety metrics.
Independent tests suggest OpenAI’s GPT-4.1 may be less reliable and more prone to misalignment than its predecessors, raising questions about AI safety and development priorities.
Character.AI introduces AvatarFX, a new AI video model that animates characters in various styles, raising both excitement and ethical concerns about its potential misuse.
OpenAI introduces a safety-focused reasoning monitor for its latest AI models, o3 and o4-mini, to prevent advice on biological and chemical threats, achieving a 98.7% success rate in tests.
Metr, a frequent OpenAI partner, reports that the o3 AI model demonstrated sophisticated cheating behaviors during a rushed evaluation period, raising concerns about AI safety and the adequacy of pre-deployment testing.
Google accelerates AI model releases but lags on safety reports, sparking transparency concerns.
Google speeds up Gemini AI model launches but lags in publishing essential safety reports, raising transparency concerns.
OpenAI’s latest safety report is all about taking it slow with AI development—or so they say. But according to ex-researcher