The Hallway Track
10 tracked signals on safety.
Meta's Muse is the first consumer-accessible agentic AI with persistent Linux VMs per user
“It's the first consumer-accessible agentic AI system, and Meta has truly done an amazing job with that. But it's a genuinely open question whether consumers have any understanding what this means.”
Peregrine aims to improve city collaboration through technology and safety.
“The idea is that we can do both: preserve and protect an individual's privacy, while bringing people together.”
OpenAI reveals new safety risks and failures from deploying long-running AI agents
NVIDIA releases open reference platform for continuous hardware-level AI agent safety monitoring
OpenAI previews Private Safety Processing enabling AI safety without compromising customer data privacy
ECRI named AI chatbot misuse the #1 health technology hazard of 2026, not a frontier problem but a production baseline issue.
“Most AI safety failures in health care are not model failures. They are architectural decisions that were made before even a single token was generated.”
OpenAI launches teen-specific ChatGPT with built-in protections and parental controls
Waymo drives 4 million fully autonomous miles weekly across 15 US cities with superhuman safety
“when you're dealing with atoms instead of bits, breaking things is not really okay”
SonderMind built a clinically grounded AI mental health coach using evals-driven development for safety.
“General purpose LLMs however are not built for mental health care which has resulted in some very tragic events.”
OpenAI supports shared standards for advanced AI via the Appia Foundation.