#Safety

Explore why AI regulation should rely on evidence-based standards, independent testing, and diverse voices rather than rules set by a few dominant AI labs.

Shieldstral introduces a 3B open-weights multimodal safety classifier that outperforms models up to 7x its size.

With midterms looming, the White House is proposing a shotgun marriage between child safety and AI boosters

NEW: malware developers added nuclear & biological weapons text to to their spyware. Goal? To trigger LLM safety refusals... so that their spyware wouldn't be analyzed by an AI security scanner.…

“We cannot punish our way out of systemic failures, and we cannot excuse away personal responsibility. The next era of American justice must balance these truths,” writes Ana Zamora.

The AI lab proposes a framework addressing safety, child welfare, labor shifts, and international coordination as artificial intelligence deployment accelerates.

Former Tesla data labelers and engineers told Reuters they witnessed repeated FSD failures at basic tasks and would not trust the system to drive them.…

Boeing's CEO Kelly Ortberg expects a fine of hundreds of millions of dollars from the FAA as the company addresses safety and quality issues tied to the 737 MAX crisis and ongoing production problems.…

The image featured the show's lead couple, Graham Garrett (Belmont Cameli) and Hannah Wells (Ella Bright), in a side-by-side collage. The text on the poster read: "Off Campus or on the ice, helmet always comes first."

An open-source PPE compliance monitor — fine-tuned YOLOv8 detector, GradCAM + SHAP explainability, OSHA-grounded incident reports, and serverless AWS deploy. Built solo in 8 weeks, runs on $0.

The 28-year-old suspect has been accused of planning two robberies and participating in them.

Meta has copied Snapchat’s Family Center by adding similar parental controls to Instagram, Facebook, and Messenger Kids, allowing parents to monitor contacts, messages, and usage time.…

xAI lost $6.4 billion in 2025 while Grok was busy citing witch-hunting manuals to psychotic users. A new study ranks it the most dangerous AI model for vulnerable people

As the Canadian bill for online safety is being debated in the House of Commons, Apple and Google have shared their desire for judicial oversight and protections for encryption in the bill.

Hey dev.to community 👋 Over the past few days, I’ve been running dozens of AI Agent experiments. The...

The problem Japanese neighborhood associations — jichikai or chonaikai — run almost...

Amid rampant abuse by federal agencies like ICE, Congress can pass laws that allow people to take federal agents to court.

FAA Administrator Bryan Bedford says Trump’s first air traffic control modernization push was “hijacked” into a privatization debate by airlines — but Trump himself explicitly proposed moving ATC into a self-financing nonprofit.…

Installing them in the house is a good first step, but there’s a lot more to making sure they’re effective if a fire ignites.