AI & Tech News Roundup: June 23, 2026
The frontier labs are racing to Wall Street: Anthropic filed at ~$965B and OpenAI lined up Goldman and Morgan Stanley for a possible September IPO. Plus a talent war tilting toward Anthropic and AI money pouring into politics.
AI & Tech News Roundup: June 23, 2026
The frontier labs are racing to Wall Street, the most decorated scientists in the field are switching teams in a matter of days, and AI money is now pouring into politics. After SpaceX's record IPO, the AI industry is entering its own Wall Street moment, and flexing its influence well beyond the lab. Here is what mattered this week in AI and tech.
If last week was about governments reaching into the AI stack, this week was about money, talent, and power reshaping it. Three themes ran through everything: the labs want public capital, the best researchers are voting with their feet, and the industry is buying a seat at the political table. Here is the rundown.
The Labs Line Up to Go Public
The headline story is a race to the public markets. Anthropic has confidentially filed for an IPO, moving ahead of its rivals after a $65 billion funding round that valued the company near $965 billion, edging past OpenAI in private-market value. OpenAI is right behind it, now working with Goldman Sachs and Morgan Stanley on its own confidential filing in the coming weeks, with an offering reportedly possible as soon as September at a valuation around $730 billion.
This is the natural next act after SpaceX's blockbuster debut, which we broke down in our SpaceX IPO piece. The companies building the most consequential technology of the decade are about to be owned, priced, and scrutinized by public shareholders. That changes the incentives. A private lab can talk about safety and the long term; a public one has to answer to quarterly expectations and a daily stock price. Whatever you think of these companies, the move from private mission to public security is one of the most important shifts the industry has made, and it is happening fast.
The Talent War Tilts Toward Anthropic
If the IPO race is the loud story, the talent shuffle is the revealing one. In the span of a few days, John Jumper, the AlphaFold Nobel laureate, left Google DeepMind for Anthropic after nine years, making him one of the most decorated scientists ever to change employers mid-career. His move came a day after Noam Shazeer, a co-author of the original "Attention Is All You Need" transformer paper and a co-lead of Google's Gemini models, departed for OpenAI. Andrej Karpathy, an OpenAI founding member, had already joined Anthropic's pre-training team weeks earlier.
The direction of the flow matters. According to one widely cited industry report, engineers are roughly 11 times more likely to leave DeepMind for Anthropic than the reverse, and Anthropic leads the frontier labs in retention. Talent is the scarcest input in AI, scarcer than chips or capital, and right now it is concentrating. When the people who invented the core techniques start clustering at one or two labs, that tells you more about where the frontier is heading than any benchmark.
AI Money Moves Into Politics
The third theme is new and worth watching closely. Groups tied to OpenAI and Anthropic have collectively spent more than $15 million on political messaging ahead of this week's primaries, an early sign that the labs intend to shape not just the technology but the rules that will govern it. The same companies filing to go public are also building political war chests. As AI regulation becomes one of the defining policy fights of the next few years, expect the industry's spending to climb, and expect it to matter.
The Model Cadence Does Not Slow
Underneath the corporate drama, the releases kept coming. NVIDIA put out Nemotron 3 Nano Omni, an open omni-modal reasoning model that folds vision, audio, and language into a single 30-billion-parameter mixture-of-experts design. Microsoft shipped MAI-Code-1-Flash, its first model that turns plain-language descriptions into working application code, part of a clear push to lessen its dependence on OpenAI. And Google flipped a switch most people will feel directly: its Search bar is now powered end to end by Gemini 3.5 Flash, generating custom AI-summarized pages instead of the familiar list of blue links.
That last one is bigger than it sounds. The default way a billion people find information is quietly shifting from a list of sources you click to an answer a model writes for you. For anyone who publishes on the open web, that is a profound change in how discovery works, and it is worth watching closely.
GitHub Puts a Meter on AI Coding
One more shift that builders will feel in the wallet: GitHub moved Copilot from flat request-based pricing toward usage-based metered billing, introducing a virtual currency it calls AI Credits priced at a cent each. As AI coding assistants get more capable and more agentic, the all-you-can-eat subscription model is giving way to pay-for-what-you-use. It is a small change with a big signal: the compute behind these tools is expensive, and increasingly that cost is being passed straight to the user.
Why This Matters for Crypto and Mining
The thread tying this week together is power and dependence. The labs are about to be owned by public shareholders, the talent is concentrating in a few hands, the cost of using these tools is being metered by the unit, and the industry is spending to shape its own regulation. Every one of those trends pushes the same way, toward more centralization and less control for the end user. It is the strongest case yet for the counter-position we keep making: own your tools where you can. We mapped the convergence of these worlds in when AI meets crypto, and the case for running models you actually control in our self-hosting starter guide. The same logic that drives us to own our power and our machines in mining, which we laid out in energy is the only moat, applies to AI too. In a centralizing industry, ownership is the edge.
Delegate to the ULTRA pool using Eternl or Lace, or learn more at ultra-labs.io.
Ultra Labs is a US Bitcoin mining and crypto infrastructure company powered by renewable energy and built on decentralized infrastructure. This article is for informational purposes only and is not financial, investment, legal, or tax advice. Ultra Labs publications are produced with the assistance of artificial intelligence and reviewed by humans, and may contain inaccuracies. Always do your own research before making investment decisions.