← Kilroy’s Daily Briefings
🧠 AI News PM

AI News Afternoon Briefing — Sunday, May 24, 2026 at 4:13 PM

🧠 AI News PM5/24/2026🕐 3:00 PM⏱ 7:10AudioPM edition

Top stories, ranked by relevance.

Story cards stay below the sticky dock while audio, chapters, date, and brief navigation remain accessible.

▶ Listen at 0:29

#1OpenAI Model Autonomously Disproves 80-Year-Old Erdos Geometry Conjecture

An OpenAI general-purpose reasoning model autonomously disproved the Erdos unit distance conjecture, an open problem in discrete geometry since 1946. The model found a proof using algebraic number theory tools never previously applied to this problem, connecting two unrelated mathematical fields. Fields medalist Tim Gowers called it "a milestone in AI mathematics," and the result has been independently verified by Princeton's Noga Alon and other leading mathematicians.

🔗 OpenAI

#2Anthropic Nears $900B Valuation, Reports First Profitable Quarter, Eyes IPO

Anthropic is closing a $30B+ funding round at a valuation above $900 billion — surpassing OpenAI's $852B to become the most valuable private AI startup. The round is co-led by Sequoia, Dragoneer, Altimeter, and Greenoaks. The company is on track for $10.9B in Q2 revenue (annualized run rate over $44B), which would mark its first-ever profitable quarter. An October 2026 IPO is reportedly under active consideration.

🔗 CNBC

#3Anthropic Signs $1.25B/Month Compute Deal with xAI for Colossus Data Center

Anthropic secured exclusive access to xAI's entire Colossus 1 data center in Tennessee — over 220,000 NVIDIA GPUs and 300+ megawatts of compute — at $1.25 billion per month through May 2029, potentially generating $40B+ for xAI. Anthropic also expressed interest in partnering with SpaceX on orbital compute infrastructure. Musk said he was "impressed" by Anthropic's team, a notable tone shift from previous criticism.

🔗 TechCrunch

#4Trump White House Kills AI Safety Executive Order After Tech CEO Lobbying

The White House scrapped an executive order that would have established a voluntary 90-day federal review process for frontier AI models before release. Eleventh-hour phone calls from Elon Musk, Mark Zuckerberg, and former AI czar David Sacks convinced Trump to pull the order hours before its scheduled signing. Trump said the order "gets in the way" of America's AI lead.

🔗 Axios

#5Google I/O 2026: Gemini Becomes a Cross-Platform Intelligence Layer, 3.5 Flash Goes GA

Google unveiled Gemini as a unified AI layer integrated across Search, Android, Chrome, Workspace, and YouTube. Gemini Spark, a 24/7 personal AI agent, will get MCP support for third-party apps (Canva, Instacart, OpenTable) within weeks. Gemini 3.5 Flash is now generally available with frontier-level performance at 4x the speed of comparable models, priced at $1.50/$9 per million tokens with a 1M context window.

🔗 CNBC

#6Leaked Audio: Zuckerberg Admits Meta Tracked Employees to Train AI Before 8,000 Layoffs

A leaked recording from a Meta all-hands meeting reveals Mark Zuckerberg confirmed Meta has been monitoring employee activity across Gmail, GChat, VSCode, and internal tools to train AI models. The mandatory tracking program logs keystrokes, mouse movements, and screen snapshots with no opt-out. The audio surfaced the same day 8,000 employees received layoff notices. Zuckerberg justified the program by saying Meta employees are "significantly" smarter than outside data-labeling contractors.

🔗 Common Dreams

#7Anthropic and Gates Foundation Launch $200M AI Partnership for Global Health and Education

Anthropic and the Gates Foundation announced a four-year, $200 million partnership — the largest deal of its kind between an AI company and a global philanthropy (4x OpenAI's $50M Gates deal). Funding covers Claude credits, grants, and technical support for programs targeting neglected diseases (polio, HPV, eclampsia), K-12 tutoring in underserved regions, and smallholder agriculture productivity improvements.

🔗 Anthropic

#8OpenAI Launches $4B DeployCo Consulting Subsidiary

OpenAI launched the OpenAI Deployment Company, a majority-owned subsidiary backed by $4B+ from TPG, Advent, Bain Capital, Brookfield, SoftBank, Goldman Sachs, and others. DeployCo places specialized AI engineers inside client organizations to build production AI systems. OpenAI simultaneously acquired applied-AI firm Tomoro, bringing 150 forward-deployed engineers on board from day one. It's a direct play at the $375B enterprise consulting market.

🔗 OpenAI

#9Ineffable Intelligence Raises $1.1B in Europe's Largest Seed Round for RL-Based "Superlearner"

Founded by former DeepMind RL lead David Silver, Ineffable Intelligence closed a $1.1B seed round at a $5.1B valuation — the largest seed in European history. Co-led by Sequoia and Lightspeed, with participation from NVIDIA, Google, and the UK Sovereign AI Fund. The company is building a "superlearner" that uses reinforcement learning self-play to discover knowledge without relying on human-generated data.

🔗 TechCrunch

#10OpenAI Launches Self-Serve ChatGPT Ads Manager, Eyes $100B Ad Business by 2030

OpenAI rolled out ChatGPT Ads Manager in beta to all U.S. businesses on May 5, with CPC bidding, no minimum spend, and measurement tools. Agency partners include Dentsu, Omnicom, Publicis, and WPP. OpenAI is targeting $2.5B in ad revenue this year, with a stated goal of $100B annually by 2030. Ads appear inside ChatGPT conversations but OpenAI says no conversation data is shared with advertisers.

🔗 OpenAI
Kilroy’s Daily Briefings · 5/24/2026