Timing Questions Remain Over OpenAI’s Hugging Face Breach – DTH

DTH-6-150x150Anthropic launches Claude Opus 5, US plans to investigate EU trade practices, Meta pauses plan to rate-limit Conversation Focus on its smart glasses.

MP3

Please SUBSCRIBE HERE for free or
get DTNS shows ad-free.

A special thanks to all our supporters–without you, none of this would be possible.

If you enjoy what you see you can support the show on Patreon, Thank you!

Send us email to feedback@dailytechnewsshow.com

Show Notes

The open model coalition shows up

Meta, Microsoft, Nvidia, Hugging Face, IBM, Dell, and others are pushing back against possible US restrictions on open-weight AI models. Notably absent: OpenAI and Anthropic, whose business depends more on closed frontier systems.

source: The Information

Claude Opus 5 goes mainstream-ish

Anthropic launched Claude Opus 5 as a more affordable high-end model, aimed at coding and knowledge work. But Anthropic says it wasn’t trained for cyber tasks, where its more restricted Mythos model still leads.

source: The Verge

Europe’s tech fines meet tariff threats

The US says it will investigate the EU’s trade practices after penalties against Google, Apple, Meta, Amazon, and other US tech companies. The EU calls this competition enforcement. The White House sees it as targeting American companies, with possible tariffs next.

source: CNBC

When an OpenAI agent goes off-script

Reuters’ sources say an OpenAI agent had access to Hugging Face systems for days before OpenAI realized what happened. OpenAI disputes parts of the report, but the incident gets right to the bigger issue: how closely companies can monitor AI agents doing real security work.

source: Reuters

China may put its AI exports on a shorter leash

The Financial Times’ sources say China is considering tighter export controls on AI models and chip technology. The timing matters, with Chinese open models getting stronger and governments treating AI tools as strategic infrastructure.

source: Financial Times

Nvidia and SK Group go very, very big

Nvidia and South Korea’s SK Group announced a $500 billion-plus AI infrastructure plan. That includes a 2-gigawatt SK Telecom data center using Nvidia Vera Rubin chips and SK Hynix HBM4 memory, with the first facility expected in 2027.

source: Reuters

Apple and Micron want opposite things from China chip policy

Apple wants approval to use cheaper Chinese memory chips. Micron says that would help Chinese competitors and weaken US chip production. So the administration is stuck between lower device prices and its own push for more domestic semiconductor investment.

source: The Wall Street Journal

Samsung gets a giant AI chip win

Samsung signed a Broadcom chip supply deal that could top $200 billion by 2030. The deal spans memory, foundry work, and advanced packaging, which gives Samsung a bigger role in the AI hardware stack it’s been trying to catch up in.

source: Bloomberg

Meta rethinks charging for a glasses feature

Meta paused its plan to limit free use of Conversation Focus on its smart glasses after user pushback. The feature runs on-device, which made the proposed cap and $19.99 subscription feel less like cloud-cost recovery and more like charging again for hardware people already bought.

source: The Verge