Gen and Intel push on device artificial intelligence deepfake detection

Cyber safety company Gen is partnering with Intel to bring on device artificial intelligence deepfake detection to consumer hardware, targeting scams that hide inside long form video and synthetic audio. New research from Gen suggests most deepfake enabled fraud now emerges during extended viewing sessions rather than through obvious phishing links.

AMD teases Ryzen Artificial Intelligence PRO 400 desktop APU for AM5

AMD has quietly revealed its Ryzen Artificial Intelligence PRO 400 desktop APU design during a Lenovo Tech World presentation, signaling a shift away from legacy desktop APU branding. The socketed AM5 part is built on 4 nm ‘Gorgon Point’ silicon and targets next generation Artificial Intelligence enhanced desktops.

Inside the new biology of vast artificial intelligence language models

Researchers at OpenAI, Anthropic, and Google DeepMind are dissecting large language models with techniques borrowed from biology and neuroscience to understand their strange inner workings and risks. Their early findings reveal city-size systems with fragmented “personalities,” emergent misbehavior, and new ways to monitor and constrain what these models do.