Sony clarifies AI Camera Assistant after viral demo – what changed and what to watch
A viral demo pushed Sony to say its Xperia AI Camera Assistant makes shooting suggestions – not secret edits. The episode highlights transparency, provenance and UX as the next battleground.
Thinking Machines wants an AI that actually listens while it talks
Thinking Machines is pitching a shift from turn-taking LLMs to streaming ‘listen-while-talk’ models. The move is technically plausible now – and it raises new privacy, latency and governance exposure.
Poppy debuts a proactive AI assistant – privacy risks to watch
Poppy launched a connector-driven assistant that continuously monitors calendar, email and messaging to surface reminders and suggested actions. The shift to background orchestration creates new privacy and governance exposure for users and companies.
YouTube expands likeness (deepfake) detection to all adult users
YouTube has opened its AI-powered likeness detection to all users aged 18+, letting people submit a selfie-style scan so the platform can search for videos that use their likeness and notify them on matches. The move raises immediate privacy, accuracy and governance questions.
Microsoft Edge lets Copilot read and synthesize your open tabs – and that raises new exposure
The Verge reports that Microsoft Edge’s Copilot can now read and synthesize the contents of all open tabs to answer questions, compare items, and summarize sessions – a convenience that increases privacy and governance risk for users and organizations.
Andon Labs’ AI radio shows why AI can’t be trusted alone
Andon Labs wired four live, unsupervised radio channels to popular LLMs. The public stress test surfaced hallucinations, moderation gaps, and brand-safety exposure for voice AI.
Runway’s video-first bet: a credible challenge to Google in AI
Runway is pivoting from creator tools to video-grounded world models. That shift is a market signal: modality-specific data moats and creator distribution could upset text-first incumbents.
AI language models threaten research integrity – a timely risk warning
New AI-assisted writing and citation tools are polishing weak or recycled papers into publications that pass peer review and then gain citations – amplifying questionable science.
Google Android 17: 9 AI-first features that raise new governance risks
Android 17 bundles nine headline features-many explicitly AI-driven-that move inference and UX hooks into the OS. That shift accelerates adoption but creates new governance, privacy and competition exposures.
Clio hits $500M ARR as Anthropic ramps LLM push – governance is the gap
Clio crossing $500M ARR and Anthropic’s platform moves mark a turning point: legal SaaS plus advanced LLMs are moving from pilots into revenue-driving production – while governance and liability lag behind.