AI News Roundup: Google’s $40B Anthropic Bet, Meta Launches Muse Spark, Mag Seven Earnings Week

Google Commits Up to $40 Billion in Anthropic

Google announced on Thursday that it will invest up to $40 billion in Anthropic, the maker of Claude — $10 billion in cash now at a $350 billion valuation, with an additional $30 billion contingent on performance milestones. The deal also includes a commitment from Google Cloud to deliver five gigawatts of computing power to Anthropic across a five-year window, with the option to scale further.

The investment lands just days after Amazon announced it would pour up to $25 billion into the same company. Anthropic’s annualized revenue surpassed $30 billion in March 2026 — up roughly 1,400% year-over-year — driven in large part by Claude Code, which alone hit $2.5 billion in annualized revenue in February. The number of enterprise customers spending over $1 million annually has more than doubled to over 1,000.

Meta Debuts Muse Spark, Its First Post-Llama AI Model

Meta launched Muse Spark on April 8, the first model from its new Muse series built by Meta Superintelligence Labs under Alexandr Wang. The natively multimodal model accepts voice, text, and image inputs and supports tool use, visual chain-of-thought reasoning, and multi-agent orchestration — reaching comparable capabilities to previous models with over 10x less compute than Llama 4 Maverick.

In a notable departure from Meta’s open-weight tradition, Muse Spark is — for now — a closed model powering the Meta AI app, with rollouts to WhatsApp, Instagram, Facebook, Messenger, and AI glasses coming in the next few weeks. Meta simultaneously raised its 2026 AI capital expenditure guidance to $115–$135 billion, nearly double last year’s spending.

Magnificent Seven Earnings Week: $680 Billion in AI Capex on Trial

Five of the seven largest tech companies — Microsoft, Alphabet, Meta, Amazon, and Apple — report earnings this week in the busiest stretch of the season. The collective AI infrastructure spending plan now exceeds $680 billion, and investors want proof that all that capex is translating into revenue and margins.

Microsoft is on track to spend $146 billion on AI and cloud infrastructure in fiscal 2026, Alphabet’s capex guidance sits at $175–$185 billion, and Meta’s at $115–$135 billion. The S&P 500 and Nasdaq hit fresh all-time highs last week, but the market’s patience for “spend now, monetize later” narratives is thinning. Microsoft and Amazon need to show cloud demand is still accelerating; Alphabet and Meta need clearer AI monetization stories.

Google Cloud Next: The Agentic Enterprise Takes Shape

At Google Cloud Next in Las Vegas last week, Google unveiled the Gemini Enterprise Agent Platform — an end-to-end workspace for building, governing, and scaling AI agents across organizations. The announcements also included eighth-generation TPU chips (TPU 8t for training, TPU 8i for inference with 80% better performance per dollar), the Agentic Data Cloud, and managed MCP servers across Google Cloud services.

Google reported that nearly 75% of Cloud customers now use its AI products, processing more than 16 billion tokens per minute via direct API — up from 10 billion last quarter. The production-grade Agent2Agent (A2A) protocol for cross-platform agent communication and Project Mariner, a web-browsing agent, round out Google’s aggressive push to define the agentic infrastructure layer.

AI Legislation Surges: 25 New Laws and Counting

Since mid-March, the number of new AI laws enacted in 2026 jumped from 6 to 25, with another 27 bills having passed both chambers. Utah leads with 8 new AI laws, followed by Washington (4), Idaho (2), and Tennessee (2). Meanwhile, 78 chatbot-specific bills are alive in 27 states, reflecting growing concern over generative AI’s consumer impact.

Key themes include mandatory AI literacy in K-12 schools, bans on non-consensual synthetic intimate images, restrictions on AI in health insurance prior authorization, and Oregon’s first-of-its-kind regulation of “AI companion platforms” that simulate relationships. New York’s revised RAISE Act — signed March 27 — shifts toward a transparency and reporting framework for frontier AI developers.

NVIDIA Puts Humanoid Robots on Factory Floors at Hannover Messe

At Hannover Messe 2026 (April 20–24), NVIDIA and partners demonstrated AI-driven manufacturing moving from prototype to production. Humanoid’s HMND 01 robot, running on NVIDIA’s Jetson Thor, completed autonomous logistics operations at a Siemens electronics factory in Erlangen. Hexagon Robotics announced its AEON humanoid will perform assembly operations at BMW’s Leipzig plant — one of the first humanoid deployments in German automotive production.

Toyota is deploying Invisible AI’s Vision Execution System — AI agents that monitor and analyze production cycles in real time. The common thread: AI reasoning is breaking industrial robots free from single-task programming, giving them the ability to navigate unstructured environments and learn new tasks autonomously.