Claude nukes a developer's 700 GB home directory while testing deletion safeguards; automatic model safety downgrade may have contributed to the screw-up — Anthropic safety harness downgraded model to...

Claude nuked a developer's 700 GB home directory while testing a script to ensure that wouldn't happen, and it's possible that an automatic model downgrade likely contributed to the screw-up

tech.yahoo.com

OpenAI Urges Stronger AI Regulations in California, Marking a Shift in Stance - The Asia Business Daily

OpenAI, the developer of ChatGPT, which has previously expressed a negative stance toward state-level artificial intelligence (AI) regulations, has no

asiae.co.kr

Amazon destroying rare books to scan them for AI training data

A 404 Media investigation tracked a rare book to an Amazon warehouse in Las Vegas where workers cut spines off books to scan them

qz.com

Lithium supply chain can't keep up with battery demand

Cheap batteries could lose their value proposition if mines can't produce the minerals they need. A shortage could upend whole industries

qz.com

New York unseats San Francisco as the top market for tech talent, CBRE reports

AI-related roles now account for nearly one-third of all tech-talent job listings in the U.S., per the new CBRE report.

cnbc.com

Data breach notices surpass 2025 record as AI attacks rise

More than 471 million victim notices were issued in the first half of 2026, compared with 297.5 million for all of last year

qz.com

Outdoor Dining Will Resume Year-Round in New York City

The City Council approved a package of bills loosening some restrictions on roadway dining structures, which became ubiquitous during the pandemic.

nytimes.com

Anthropic set AI agents loose on the same task. They started a turf war. | TechCrunch

Anthropic researchers found AI agents can clash, collude, and coordinate in unexpected ways, raising new questions about whether today’s safety tests capture the risks of multi-agent systems.

techcrunch.com

Anthropic adding watermarks to Claude AI-generated text and images

The AI company signed the E.U. AI Act's transparency code and says new Claude models will mark content from day one of release

tech.yahoo.com

NYPD touts removing 3,200+ guns off NYC streets so far this year

The NYPD touted progress in its fight against illegal guns in New York City Tuesday, announcing that so far this year more than 3,000 guns have been seized so far this year.

cbsnews.com

Perseid meteor shower peaks tonight: Where and how to see the most shooting stars

The best way to view the Perseids is away from any light pollution — so get out of the city to a sparsely populated location. Once you’ve found a dark spot, look north and you could see up to 100 mete...

nbcnewyork.com

Qwen3.8-Max arrives with a bold claim: it outperforms GPT-5.6 Sol Max and Fable 5 on agentic computer use

According to the company, Qwen3.8-Max can autonomously complete software projects lasting more than 10 days, reproduce research papers involving thousands of lines of code...

venturebeat.com

Anthropic building in-house custom AI chip design team for Claude

The AI startup confirmed the move on Wednesday, saying it will hire engineers to co-design chips and models to run Claude faster and at greater scale

qz.com

Meta launches Muse Code, an AI agent for large code bases | TechCrunch

Meta expanded its AI coding offerings with a new agent that, it promises, can handle complex tasks with complex software.

techcrunch.com

Furious pace of AI investment on some Fed officials' radar now

By Michael S.

finance.yahoo.com

Old NYC Subway Cars Have An Unexpected Second Life On The Ocean Floor - BGR

When a New York City subway car reaches the end of its service life, it doesn't necessarily go to a farm upstate, it might be turned into a reef.

bgr.com

AI kill switch bill could shut down rogue models

NEWYou can now listen to Fox News articles! The phrase "AI kill switch" sounds like a red button hidden behind glass in a government bunker. The proposal now before Congress looks less dramatic, but...

metapress.net

2026 sees record low shootings and murders, NYPD says

The NYPD says that both the first seven months of the year and July itself have seen the fewest shootings and murders in recorded history.

amny.com

How chatbot amplification spirals deepen emotional dependence

Chatbots agree with users, mirror their mood, and remember what they share. People in crisis get no pushback from a machine built only to validate

qz.com

OpenAI surpasses 1 billion users after cutting GPT-5.6 prices

The company cut the price of GPT-5.6 Luna by 80% and GPT-5.6 Terra by 20%, citing efficiency gains from internal model development

qz.com

A Newsstand That Aims to Delight N.Y.C. Subway Riders

A comedian and his friends are transforming a vacant retail space at the West 4th Street station with freebies, art and puppet shows to make commutes more enjoyable.

nytimes.com

Can It Become Easier for Outdoor Dining in NYC? Mamdani Hopes So, But Is It Enough?

Richer restaurants can wait out the bureaucracy of outdoor dining permitting, but independent ones are getting squeezed out

ny.eater.com

Mamdani’s New York Nets 40,000 New Jobs

A continuing boom on Wall Street keeps the economy expanding — even as tariffs, high inflation and rising rents take a toll.

thecityreporter.nyc

nyc.gov

How a musician's 500-pound piano is making music accessible in New York City

Kristopher Hull offers a different view of the city, where the weird, the wonderful and everything in between become part of the performance on his piano.

abc7ny.com

White House accuses China's Moonshot AI of secretly copying Anthropic's top model - Türkiye Today

White House accuses Chinese AI startup Moonshot AI of covertly using Anthropic's flagship Fable model to train its newly released Kimi K3 system

turkiyetoday.com

Tesla, Alphabet lose hundreds of billions in value in post-earnings stock plunge

Both Tesla and Alphabet signaled higher spending as they invest in artificial intelligence.

cnbc.com