25 firms defend open weights – Anthropic, OpenAI abstain
The fight over open-weight AI entered a new round over the last week.
On July 22, White House officials accused China’s Moonshot AI of stealing American IP through distillation and floated sanctions and repercussions. Anthropic’s policy chief called the alleged operation “industrial espionage.” Then the startups showed up, with nearly 200 co-signing a letter to President Donald Trump urging him not to cut off the open models they rely on. On Friday, a statement signed by 25 major organizations warned Washington against “premature restrictions” on open weights. And Nvidia’s Jensen Huang, asked by Axios whether open labs should be allowed to distill closed models, answered that distillation is “fundamental to intelligence.”
Let me be clear: Running an open-weight Chinese AI model is not a fringe act, and it’s not disloyal. It’s how many AI-native startups and developers keep their token bills under control because Anthropic and OpenAI are becoming too expensive to build on at scale.
Washington wants to draw the line between legitimate openness and Chinese theft. Inside the technology industry, the more revealing line is open vs. closed. Microsoft, Nvidia, Meta, and 22 other organizations signed the statement, warning against premature restrictions and defending distillation as a legitimate model-development technique. Anthropic and OpenAI did not sign.
For developers, the stakes are not abstract. The affordable model they can download, adapt, and run themselves is exactly what this fight could make harder to access.
“We love the world where we can use both”: How Nvidia thinks about local and frontier models
Joey Conway, Nvidia’s senior director of generative AI software, spoke to The New Stack about how local and open models are increasingly working alongside frontier models, often with a router in between to decide which one to use, and how organizations can adapt these open models to their own needs.
Most access tools were built for humans, not for CI/CD jobs, service accounts, or AI agents making tool calls across your infrastructure. That mismatch is forcing teams to stitch together separate systems for every type of identity. Join us live on July 28 to see what a single, unified access architecture actually looks like today, without adding IT overhead or slowing your engineering teams down.
Your agents are hallucinating. Learn how to fix that.
Logs, metrics, and traces work great until an AI agent starts hallucinating or overconsuming. In this session, Dotan Horovits (CNCF Ambassador) and Rekha Thottan (AWS) walk through how to monitor traditional services and evaluate AI agent behavior in production, using a single open-source stack built on OpenSearch and OpenTelemetry.
The #1 mistake SecOps teams make with AI, according to an ex-NSA red teamer.
Chas Clawson ran Red Team assessments at the NSA before architecting the SIEM stack for a federal security operations center. In this insightful webinar, he’s telling us what that taught him about AI, alerts, and where most SecOps teams still get it wrong.
WeAreDevelopers brings the world’s largest developer event to North America, fostering global growth and connection in the tech sector. Join The New Stack in San Jose for all the action as we host a special welcome reception, cover the biggest show moments, and film interviews on the show floor at our booth. We look forward to seeing you there! Use this exclusive code to save 10% on your registration: thenewstack_community
Google shares a single compute base across its fleet of services, including bringing its AI products to bear in search, YouTube, and its enterprise products.
Google (Alphabet) sells compute and inference to third parties and must apportion compute for both its internal use and the training of new models.
I wonder if the company is more constrained internally than we might think?
Could Google’s recently slow AI progress indicate that it is making painful tradeoffs that may impact its ability to regain the frontier?
.png?height=72&upscale=true&name=The%20New%20Stack%20logo%20-%20Reversed%20(2).png){kind=small}
.png?upscale=true&width=1168&upscale=true&name=social_LiveNow_vs-2_16x9%20(4).png)
