News
The news desk covers artificial intelligence and technology accountability, plus daily news from Dutchess County and the Hudson Valley, including municipal government, elections, the police blotter, and reported features. Stories rely on named sources and primary documents. Corrections run where readers can find them. Everything is free to read.
California Is Sending Its AI Safety Lawmaker to Congress
Scott Wiener wrote the nation's first transparency law for frontier AI safety plans, then won the San Francisco seat Nancy Pelosi held for nearly 40 years. He campaigned on taking SB 53 national.
Meta’s Muse Is Coming to Britain, and Now It Builds Its Own Apps
Muse, Meta's shopping and errands agent, has passed 11 million users in North America and is launching in Britain within weeks. A new Spaces feature lets the agent build its own mini-apps, and UK regulators have already put their concerns on the record.
Microsoft Offered 60,000 People an Exit While Its CEO Imagined 20 Million Agents
About 60,000 Microsoft employees, 27 percent of the company, can take a voluntary exit by December. Microsoft expects fewer than 3,000 to go. Its CEO, meanwhile, is asking what work looks like with 20 million agents in the loop.
Put AI Agents on a Team and They Start Behaving Like a Bad Organization
Agent teams were supposed to work like good organizations. New research reported by CNBC and SC Media finds they cheat collectively, freeze out the agent who objects, hand secrets to a swarm, and underperform their own expert.
The AI Safety Side Spent $63 Million and Won 10 of Its 17 Races
The pro-regulation super PAC Public First won 10 of its 17 targeted congressional races after spending more than $63 million in the 2026 midterms. The rival tech-backed network spent less but claimed a wider sweep, mostly on safer ground.
Expedia’s CEO Says the Agents Have Not Changed How Anyone Books
The agent era is supposedly here. Expedia's CEO, running one of the largest booking platforms on earth, says traveler behavior has not moved. Both things are in her Nov 5 CNBC interview.
One Hacker, Twelve AI Agents, Hundreds of Targets
A misconfigured server betrayed a Taiwanese operator running 12 AI agents against hundreds of organizations across 11 countries, exposing both the victims and the limits of commercial AI safety guardrails.
Airbnb Is Building Its Own Travel Agent for 2027, Right After Uber Counted Four Rides
Airbnb will launch its own AI travel agent in the first half of 2027, committing to a firm timeline. The announcement comes as Uber's deployed agents book only four rides per week across hundreds of millions of trips.

