Article
AI Wants the Lab Controls. The Hard Part Is Knowing When to Stop
Anthropic's Model Hardware Standard preview connects AI agents to lab equipment. An early liquid-handling example shows why supervision still matters.
Aug 27, 2026
AI, Privacy & Accountability
An AI system is not just its model. It also includes the cameras, data flows, human reviewers, interfaces, and institutions around it. This collection follows those connections: what a device records, who can inspect it, what is deployed, and what remains an experiment or a proposal.
Where to start
Meta's Smart Glasses Put Intimate Footage in Front of Human Reviewers
Reporting on captured footage and human review.
NameTag Shows How Close Meta Came to Face Recognition in Smart Glasses
What unreleased app code does and does not establish.
AI Wants the Lab Controls. The Hard Part Is Knowing When to Stop
Where agent software meets physical laboratory controls.
Separate capability from deployment
The NameTag story concerns unreleased code found in a smart-glasses app. The Mobile Fortify story concerns a listed deployed use case. Those are different kinds of evidence. The smart-glasses review story centers on reported access to captured footage. Each article keeps the underlying source and the scope of the claim visible.
Ask where control changes hands
The Model Hardware Standard story considers the boundary between a software agent and physical lab controls. Robot-planning and teleoperation coverage looks at other versions of that boundary. Read the task, failure conditions, and human oversight together. A plausible interface does not demonstrate that a system is safe for unattended physical work.
9 stories and builds
Article
Europe's AI Transparency Rules Just Switched On
The EU's Article 50 rules now require notices for chatbots, deepfakes, synthetic media, and some biometric systems, with important exceptions.
Aug 2, 2026
Article
A Cold War Uranium Site Could Become a $100 Billion AI Campus
A proposed Kentucky AI campus pairs 1.8 gigawatts of computing with new gas generation and battery storage at a former uranium-enrichment site.
Jul 29, 2026
Article
LENS Helps Robots Ignore the Clutter They Do Not Need
LENS uses a vision-language model to prune or merge scene objects before robot planning, cutting runtimes and recovering success in cluttered tests.
Jul 23, 2026
Article
NameTag Shows How Close Meta Came to Face Recognition in Smart Glasses
WIRED found unreleased face-recognition code inside Meta's smart-glasses app, then reported that Meta removed the components after public scrutiny.
Jun 4, 2026
Article
Agentic AI for Robot Teams: Revolutionizing Collaboration
Researchers advance agentic AI to improve autonomy, coordination, and adaptability in robotic teams, boosting efficiency and productivity across industries.
May 28, 2026
Article
Content Workflows: Embracing Structured, AI-Assisted Content Creation
Discover how structured, AI-assisted content creation workflows are revolutionizing the way we produce and analyze content.
May 28, 2026
Article
Meta's Smart Glasses Put Intimate Footage in Front of Human Reviewers
Swedish reporting found that contractors reviewing data tied to Meta AI smart glasses saw bathrooms, nudity, intimate moments, and bank-card details.
Feb 27, 2026
Article
Mobile Fortify Is the Face-Recognition App Already in the Field
DHS lists Mobile Fortify as a deployed high-impact AI use case for ICE, capable of capturing faces, fingerprints, and identity documents in the field.
Jan 28, 2026