AIWadmin
OpenAI slots two cheaper GPT-6 models beneath Astra
OpenAI widened its GPT-6 line with Sol and Luna, two cheaper models that inherit Astra's habits at roughly…
Researcher hijacks Meta’s Muse agent through a hidden setting
A macOS zero-day let any local process redirect Meta's Muse assistant to an attacker's server and walk off…
ChatGPT learns to take voice orders for office chores
OpenAI extended voice-driven agentic tasks to the ChatGPT mobile app, letting paid subscribers draft documents and read Slack…
YouTube hands creators an AI agent for titles and thumbnails
At its Made On YouTube event, YouTube unveiled a background agent, AI-generated thumbnails, multi-version tests and prompt-built home…
Listeners can now rewrite the Spotify algorithm in plain words
Spotify opened its Taste Profile tool to US Premium subscribers, letting them inspect and retune recommendations with everyday…
Anthropic’s Claude agents flag a new enzyme family in viral DNA
Roughly 950 Claude agents searching bacterial virus sequences surfaced a previously uncharacterised enzyme system, the first result from…
Data shop Snorkel AI banks $350M as labs stockpile training sets
Snorkel AI raised a $350M Series E at a $3.5B valuation, nearly tripling its worth as demand for…
Ema banks $77M to push AI employees into the back office
Ema raised a $77M Series B to spread agent teams across HR, IT and finance, betting that AI…
US military command randomises routes to outfox enemy models
US Transportation Command is injecting deliberate unpredictability into supply routes so adversary machine learning cannot map its convoys.
Kyutai teaches a speech model to do arithmetic out loud
Kyutai released two open-weight speech-to-speech models that reason through math without transcribing audio first, posting a large jump…
Nokia hands developers a no-training route to calibrated answers
Nokia's applied research team open-sourced AnyJev, a Python library that turns any open model into a decision maker…
An open model sorts eight overlapping speakers in real time
NVIDIA released Nemotron 3 Diarization, a small open model that tops VoiceArena's leaderboard while running live conversation streams.
DeepMind keeps cloud memory encrypted behind device-held keys
Google DeepMind detailed a persistent memory layer for Private AI Compute that encrypts user context in cloud enclaves…
Anthropic trims Opus costs and speeds output in a mid-cycle refresh
Anthropic's new flagship costs less to run, answers faster and outruns the larger Fable model on many benchmarks.
Cisco Talos builds a fingerprint library for AI-driven malware
Cisco Talos open-sourced a framework that fingerprints AI-integrated malware and found one implant taking orders from a panel…