AI moves fast. Stay ahead.

Google unveiled Gemini 3.5 Flash and Gemini Spark — a 24/7 personal AI agent for task automation — at I/O on May 19. Gemini is now the intelligence layer across Search, Android, Workspace, and XR smart glasses, marking Google's decisive shift from chatbot to agent-first platform. TechCrunch ↗ · Google Blog ↗

Anthropic reported Q2 projected revenue of $10.9B and $559M operating profit — the first profitable quarter by a frontier AI lab. The company is raising $30B+ at a $900B valuation ahead of a planned October IPO. Anthropic also surpassed OpenAI in enterprise adoption (34.4% vs 32.3%) per Ramp's AI Index. CNBC ↗ · TechRepublic ↗

The US Commerce Department's CAISI finalised pre-deployment evaluation agreements with OpenAI, Anthropic, Google DeepMind, Microsoft, and xAI. Every major frontier model now undergoes government review before public launch — a historic first. The EU began separate talks with Anthropic about early Mythos access. Build Fast with AI ↗

A federal jury in Oakland ruled unanimously against Elon Musk in under two hours, finding his claims were filed outside the statute of limitations. The verdict cleared OpenAI's path to a potential $1 trillion IPO. OpenAI's S-1 revealed Q1 revenue of $818M but a $2.47B operating loss. CNBC ↗ · Yahoo Finance ↗

An OpenAI reasoning model disproved the planar unit distance conjecture posed by Paul Erdős in 1946 — the first AI-generated mathematical proof that Fields Medallist Tim Gowers said he would recommend for publication "without any hesitation." The model used algebraic number theory, not classical geometry. Scientific American ↗ · TechCrunch ↗

Figure's Helix-02 humanoid robots ran a live multi-hour autonomous warehouse operation, sorting thousands of packages per shift with no human intervention. The robots self-managed battery swaps, diagnosed failures, and coordinated as a fleet — one of the clearest public demonstrations of long-duration autonomous multi-robot operation. TechTimes ↗ · eWeek ↗

Meta cut 8,000 employees (~10% of workforce) and reassigned 7,000 more to AI teams while cancelling 6,000 open roles. The company simultaneously raised its 2026 AI capex guidance by $10B to $145B. Tech industry job cuts have surpassed 100,000 in 2026 so far, many attributed to AI automation. CNBC ↗ · TechSpot ↗

AMD began production of its 6th Gen EPYC Venice processors on TSMC's 2nm node — the first HPC product at this process node — challenging NVIDIA's AI compute dominance. Simultaneously, Meta signed a multi-year deal to deploy up to 6 gigawatts of AMD Instinct GPUs with initial shipments targeted for H2 2026. Crescendo AI ↗ · StockTitan ↗

Pope Leo XIV released his 42,300-word encyclical on AI on May 25, with Anthropic co-founder Christopher Olah present at the Vatican ceremony. The document calls for AI to be governed by human dignity, warns against autonomous weapons, and compares AI governance to nuclear arms control. Tech figures including Jack Dorsey shared the text. CBS News ↗ · TIME ↗

The US Department of Health and Human Services announced it will use ChatGPT and other AI tools to analyse annual audit reports from all 50 states, targeting fraud, waste, and abuse in federal health spending. The programme covers programmes representing approximately $2 trillion in annual expenditure. Crescendo AI ↗