AI Briefing

Meta employees burned 60 trillion AI tokens in 30 days competing for 'Token Legend' status on an internal leaderboard called Claudeonomics... and then it got shut down because someone leaked it. Meanwhile, OpenAI is pitching Congress on robot taxes and a four-day workweek, and NVIDIA just dropped open models purpose-built for humanoid robots. Oh, and OpenAI published a child safety blueprint that reads like a wake-up call for the entire industry.

01IndustryFortune ↗

Meta Shuts Down 'Claudeonomics' After Employees Burn 60 Trillion AI Tokens Competing for Leaderboard Glory

A Meta employee built an internal dashboard called 'Claudeonomics' that ranked the company's 85,000+ employees by AI token consumption, awarding titles like 'Token Legend' and 'Cache Wizard' to the top 250 users. In just 30 days, total usage exceeded 60 trillion tokens at an estimated cost of $9 billion, with the top user consuming 281 billion tokens. The dashboard was shuttered on April 9 after data was leaked externally, with the creator taking it down voluntarily. The most revealing detail: neither CEO Mark Zuckerberg nor CTO Andrew Bosworth ranked in the top 250, raising questions about whether the people steering Meta's AI-first strategy are actually using the tools themselves.

02PolicyTechCrunch ↗

OpenAI Proposes Robot Taxes, a Public Wealth Fund, and a Four-Day Workweek in 13-Page Policy Blueprint

OpenAI published a sweeping 13-page policy document calling for structural economic reforms to prepare for what it describes as approaching superintelligence. The proposals include taxing automated labor and shifting the tax burden from workers to capital, creating a national public wealth fund seeded partly by AI companies, and government-subsidized pilots of a 32-hour workweek with no pay cuts. OpenAI warns that AI-driven productivity gains could hollow out the tax base funding Social Security, Medicaid, and housing assistance. Coming weeks before the company's expected IPO at a valuation north of $300 billion, the document reads as both genuine policy advocacy and strategic pre-IPO positioning to frame OpenAI as a responsible steward of transformative technology.

03PolicyOpenAI ↗

OpenAI Publishes Child Safety Blueprint Urging Industry-Wide Action on AI-Generated CSAM

OpenAI released a comprehensive child safety policy framework on April 8 proposing coordinated action across industry, government, and civil society to combat AI-enabled child sexual exploitation. The blueprint calls for modernizing laws to explicitly cover AI-generated and AI-altered child sexual abuse material, improving provider reporting and coordination for more effective investigations, and building safety-by-design measures directly into AI systems. OpenAI cites Internet Watch Foundation data showing over 8,000 reports of AI-generated CSAM in just the first half of 2025, a 14% year-over-year increase. The document also proposes a good-faith safe harbor for companies conducting CSAM detection, red-teaming, and reporting to NCMEC.

04Open SourceNVIDIA Newsroom ↗

NVIDIA Releases Cosmos Reason 2 and GR00T N1.6 Open Models for Humanoid Robots During National Robotics Week

NVIDIA used National Robotics Week to release two major open models for physical AI: Cosmos Reason 2, an open reasoning vision-language model for spatial understanding, and Isaac GR00T N1.6, an open reasoning vision-language-action model purpose-built for humanoid robots. The company also launched Isaac Lab-Arena for standardized robot evaluation and the OSMO edge-to-cloud framework to simplify robot training workflows. Industry heavyweights including Boston Dynamics, Caterpillar, LG Electronics, NEURA Robotics, Figure, Universal Robots, and YASKAWA are already building on the platform. The releases signal that NVIDIA is positioning itself as the default infrastructure layer for the entire robotics stack, from simulation to deployment.