AI safety center shifts toward White House & Presidential tech council meeting postponed again - Tech News (Nov 8, 2026)

One AI company is now building its systems about twice as fast as it did a year ago, and part of the reason is that the AI is helping to build itself. We'll get to that in a moment. Welcome to The Automated Daily, tech news edition. The podcast created by generative AI. I'm TrendTeller, and today is November 8th, 2026. On the docket: questions about the independence of US AI safety work, fresh numbers from the trillion-dollar funding race, new AI agents and the rules that might govern them, a legal fight over kids and social media, and a look at cybersecurity in Japan and connectivity in the Gulf. Let's get started.

AI safety center shifts toward White House

We start in Washington, where the government's main AI safety shop appears to be drifting closer to political control. Staff at the Commerce Department's Center for Advancing Innovation and Standards for Super Intelligence were reportedly asked to informally report to a White House office. The division currently has no permanent leader, and it has stopped publishing research. Former officials are uneasy. Their worry is that if evaluations of AI systems run through a political office, the results could be shaped by politics rather than evidence, weakening one of the few independent checks on powerful models.

Presidential tech council meeting postponed again

Staying in the capital, an update on the President's Council of Advisers on Science and Technology. Its scheduled meeting has been delayed once more, now to November 13. The panel counts executives like Nvidia's Jensen Huang and Meta's Mark Zuckerberg among its members, and the postponement lands in the middle of a sharp tech-stock sell-off, along with growing concern that the incoming Congress could reshape AI policy and infrastructure spending.

Anthropic doubles its AI development speed

Now to our headline item. The Model Evaluation and Threat Research group, better known as METR, reports that Anthropic is developing AI systems roughly twice as quickly as it did last year. What stands out is the reason: AI tools themselves are helping accelerate the work. That's a notable data point in the long-running debate about AI speeding up its own progress. It also creates an awkward contrast, since Anthropic's CEO Dario Amodei has been one of the loudest voices urging caution about increasingly powerful models. Building faster while warning about speed is a tension the company will likely be asked about.

OpenAI raises $11 billion first tranche

Speaking of Anthropic, there's movement on the money side. Investors have reportedly lowered their expectations for the company's eventual IPO to somewhere around one trillion to one point three trillion dollars, with some estimates sitting near the bottom of that band. The recalibration comes as the AI market sell-off pushes people to ask how sustainable these enormous private valuations really are.

Anthropic IPO valuation expectations cool

OpenAI, meanwhile, has locked in an initial eleven billion dollars of new funding at a reported valuation of one point two five trillion dollars. The company had reportedly weighed raising more than thirty billion, but trimmed the first tranche. Even scaled back, it's a reminder of just how much capital frontier AI devours, and that investors are still willing to write very large checks for the leading labs.

Thinking Machines unveils Inkling 2 Small

On the model front, Thinking Machines Lab has released Inkling 2 Small. Don't let the name fool you, it's a very large model, built so that only a fraction of it works on any given task, which keeps it efficient. It's available through the company's Tinker platform, and openly licensed weights are reportedly on the way. Open releases at this scale put serious capability into more hands and turn up the competitive heat on everyone else.

Perplexity Teammates and Cloudflare agent coalition

AI agents had a busy day too. Perplexity introduced Teammates, a feature in its Computer platform that runs persistent agents for ongoing work. These agents can take on different roles, coordinate through Slack, and run in their own isolated environments. It's another step away from simple chat and toward software that carries out multi-step jobs over time. And as agents multiply, so does the question of who they really work for. Cloudflare is tackling that with a new Trusted Agent Delegation coalition, focused on identity, reputation and accountability for AI agents. The aim is shared standards so websites can tell an authorized agent from a rogue bot. If it catches on, it could become a kind of passport system for automated visitors on the web.

Tech giants sue over California child-safety law

Over to the courts. Google, Meta and TikTok have each filed lawsuits challenging a California law designed to protect children from social-media risks. The companies argue it violates their First Amendment rights. However it plays out, the outcome could set the boundaries for how far states can go in dictating platform design and safety features for young users.

Japan launches emergency AI cybersecurity package

In Asia, Japan has announced an emergency cybersecurity package that includes AI-powered teams and specialized cyber-rescue units. The goal is to sharpen the country's ability to respond to digital attacks, reflecting a broader recognition that cyber incidents now demand faster detection, better coordination and hands-on technical help.

MWC26 Doha spotlights MENA connectivity

And finally, MWC26 Doha has opened in Qatar, with industry and government leaders calling for closer alignment on global connectivity standards. Organizers say the Middle East and North Africa mobile economy could add four hundred seventy billion dollars by 2030, putting regional infrastructure and the digital divide at the center of the conversation.

That's the roundup for November 8th, 2026. The thread running through today: AI is moving faster, costing more, and raising tougher questions about who keeps watch. Thanks for listening to The Automated Daily, tech news edition. I'm TrendTeller, and I'll see you next time.

AI safety center shifts toward White House & Presidential tech council meeting postponed again - Tech News (Nov 8, 2026)

Our Sponsors

Today's Tech News Topics