Friday, May 29 May 29, 2026
Good morning. It's Friday, May 29, 2026, and the AI industry just crossed a new threshold. Today's briefing has one unmissable center of gravity: Anthropic is now valued at nearly one trillion dollars — and they announced a major new model on the same day. But that's just the start. We've got a landmark state AI safety law, a new kind of conversational AI app built by the founders of Oculus, Apple's leaked Siri overhaul, and an enterprise AI company quietly becoming one of the most important businesses you've never heard of. Let's get into it.
---
The biggest story: Anthropic has closed a $65 billion Series H funding round at a $965 billion post-money valuation — making it the highest-valued AI startup in history, edging out OpenAI's last valuation of $730 billion. The round was co-led by Altimeter Capital, Sequoia, Dragoneer, Greenoaks, Coatue, Capital Group, and D1 Capital Partners. Infrastructure partners including Samsung, SK Hynix, and Micron also joined. A portion of the round — $15 billion — came from previously committed hyperscaler investments, including Amazon's $5 billion pledge announced in April.
For context: Anthropic's run rate revenue just crossed $47 billion. The company expects a 130 percent revenue surge to bring it to its first operating profit, according to the Wall Street Journal. With an IPO reportedly on the horizon, this is very likely the last private fundraising event before Anthropic becomes a public company. Anthropic says the capital will go toward safety research, compute expansion, and scaling the products and partnerships its enterprise customers depend on.
---
On the same day as the funding announcement, Anthropic released Claude Opus 4.8 — the newest version of its flagship model. Coming just 41 days after Opus 4.7, the rapid turnaround reflects intense competitive pressure from OpenAI Codex and Google's Gemini Flash. The headline capability: Opus 4.8 is significantly more honest about what it doesn't know. Early testers, including Bridgewater Associates, said the biggest difference was the model's tendency to proactively flag issues with inputs and outputs — something prior models routinely missed. Anthropic is calling this a major step toward reliability in high-stakes enterprise work.
Also launching alongside Opus 4.8 is a new feature called Dynamic Workflows — currently in research preview — which enables large models to orchestrate hundreds of parallel subagents. Anthropic says Claude Code can now carry out codebase-scale migrations across hundreds of thousands of lines of code, from kickoff to merge. Their most powerful model, Mythos — held back from general release over cybersecurity concerns — may be coming soon. Anthropic says, quote, "we expect to bring Mythos-class models to all our customers in the coming weeks."
---
On the regulatory front, Illinois is on the verge of enacting what may be the country's most rigorous state AI safety law. Governor JB Pritzker says he plans to sign a bill passed by the state legislature this week that would require third-party independent audits and whistleblower protections at AI companies. That goes beyond what California and New York have recently passed. If signed, Illinois becomes a new benchmark for AI governance in the United States, and it signals that state-level AI policy is moving faster than federal action.
---
In enterprise AI, Asana has acquired StackAI — a no-code agent builder — and is integrating its workflow automation technology directly into the Asana platform. The move signals that project management software is transforming into an orchestration layer for AI agents. Asana isn't alone here — every major workflow tool from Monday-dot-com to Notion is racing toward the same destination. The question is no longer whether AI agents will run inside enterprise workflows, but which platform will own the orchestration interface.
Separately, Glean — often described as the Google for the enterprise — just crossed $300 million in annualized revenue, tripling from $100 million just 15 months ago. What's driving growth? Their pitch has shifted: Glean now positions its context graph technology not just as enterprise AI search, but as a way to dramatically reduce AI token consumption costs. As AI infrastructure bills soar across corporate America, cost control is a serious selling point. Customers including Databricks, Reddit, and Pinterest are paying for efficiency, not just capability.
---
On the consumer side, two big developments in the battle for your attention.
Sesame — the conversational AI startup co-founded by the original Oculus VR founders — has launched its iOS app publicly, now available in 39 countries. What makes Sesame different: it's engineered to feel like a real conversation. The AI can run parallel searches mid-sentence, weaving in new information as it talks, the way a human pivots naturally mid-thought. The app offers four distinct AI agents — Maya, Miles, Simone, and Charlie — each with their own voice, personality, and persistent memory. More than a million people accessed the research preview in the first few weeks. The long-term roadmap includes AI-powered intelligent eyewear launching in 2027.
Meanwhile, Bloomberg has published leaked renders of Apple's upcoming Siri overhaul ahead of WWDC in June. A new standalone Siri app will compete directly with ChatGPT, Claude, and Gemini — allowing photo uploads, document analysis, and persistent chat history. Under the hood, Apple is using Google's Gemini AI. The swipe-down gesture on iPhone will now trigger an AI-powered search interface. Apple's play is scale: ChatGPT has 900 million weekly active users. Apple has 2.5 billion active devices.
---
Finally, a candid moment from inside Amazon. The company quietly shut down an internal AI usage leaderboard that was meant to encourage employees to use AI tools more. What happened? According to the Financial Times, some workers figured out how to game the rankings by assigning AI agents to run pointless tasks — a practice insiders call "tokenmaxxing." As computing gets more expensive, unnecessary AI usage at scale becomes a real cost problem. Amazon exec Dave Treadwell told employees: "Don't use AI just for the sake of using AI." A real, grounding reminder amid all the trillion-dollar headlines.
---
And that brings us to today's single business idea — the one I think this news cycle points to most clearly.
THE OPPORTUNITY: Enterprise AI Spend Auditing and Optimization as a Service.
Here's the setup. Companies are spending tens of millions on AI compute — often with little visibility into where the spend is actually going. Amazon just learned this the hard way. Glean is selling cost reduction as its top feature. And as Anthropic closes a nearly trillion-dollar round with Mythos-class models still to come, model costs are only going to increase.
What doesn't exist at scale yet: a clean, vendor-neutral AI cost intelligence layer — a SaaS product that connects to a company's cloud and AI vendor accounts, maps actual AI usage to business outcomes, and flags waste, redundancy, and misaligned consumption patterns.
Who pays: CFOs, CIOs, and IT procurement teams at mid-to-large enterprises. This is a CFO-level conversation, not just an engineering one. That means shorter sales cycles than pure platform plays, and renewal is sticky because the ROI is measurable in dollars.
Why now: The combination of skyrocketing model costs, employee tokenmaxxing, and the entrance of agents running thousands of calls per day has created a new category of AI spending sprawl. Nobody has a clear map. Glean proves customers will pay for token efficiency. The auditing version of that is a distinct, purpose-built product opportunity.
What makes it defensible: proprietary benchmarks correlating token consumption with business value — which become more accurate with each client. Over time, that data moat makes you the standard for how AI ROI gets reported.
---
That's your AI morning briefing for Friday, May 29, 2026. Anthropic is nearly a trillion-dollar company. Claude is getting more honest. Illinois is getting serious about AI safety. And the war for your enterprise AI budget is being fought on the cost control front, not just the capability front. Have a great weekend, and we'll see you Monday.
---
SHOW NOTES: • Anthropic raises $65B Series H at $965B valuation — TechCrunch (May 28, 2026) https://techcrunch.com/2026/05/28/anthropic-raises-65-billion-nears-1t-valuation-ahead-of-ipo/ • Anthropic releases Claude Opus 4.8 with new Dynamic Workflows tool — TechCrunch (May 28, 2026) https://techcrunch.com/2026/05/28/anthropic-releases-opus-4-8-with-new-dynamic-workflow-tool/ • Illinois close to enacting AI safety law with independent audits — The Verge (May 28, 2026) https://www.theverge.com/ai-artificial-intelligence • Asana acquires no-code agent-builder StackAI — TechCrunch (May 28, 2026) • Glean top line crosses $300M, AI budget-cutting becomes its major selling point — TechCrunch (May 28, 2026) https://techcrunch.com/2026/05/28/gleans-top-line-crosses-300m-as-ai-budget-cutting-becomes-its-major-selling-point/ • Sesame conversational AI app from Oculus founders launches iOS — TechCrunch (May 28, 2026) https://techcrunch.com/2026/05/28/sesame-the-conversational-ai-startup-from-oculus-founders-launches-its-ios-app/ • Sneak peek at new Siri app reveals Apple's plans to take on ChatGPT — TechCrunch (May 28, 2026) https://techcrunch.com/2026/05/28/sneak-peek-at-new-siri-app-reveals-apples-plans-to-take-on-chatgpt-and-more/ • Amazon shuts down internal AI usage leaderboard after tokenmaxxing problem — The Verge (May 28, 2026) https://www.theverge.com/ai-artificial-intelligence