Top stories in artificial intelligence — curated daily
Alphabet announced it will invest $10 billion immediately in Anthropic at a $350 billion valuation, with up to $30 billion more contingent on performance milestones. The deal also includes 5 gigawatts of compute capacity for Anthropic's model training. The announcement came just days after Amazon disclosed its own commitment of up to $25 billion in the Claude maker. Combined, the two cloud giants have now pledged as much as $65 billion to Anthropic — a bet that the frontier AI race will be dominated by a handful of well-capitalized labs.
As Google's investment was announced, Anthropic disclosed that its annual revenue run-rate has surpassed $30 billion, a more than 3× jump from roughly $9 billion at the end of 2025. The explosive growth is driven primarily by Claude Code, its agentic coding tool, which has found deep traction among enterprise developers. Anthropic's bet that coding would be the first killer use-case for frontier AI appears to be paying off — and it's one reason investors are competing to lock in early ownership.
OpenAI's latest model, GPT-5.5, arrived just six weeks after GPT-5.4. The release doubles down on agentic use-cases: multi-step task completion, computer use, code writing and debugging, and online research. GPT-5.5 now powers Codex, OpenAI's agentic coding application, running on NVIDIA GB200 NVL72 rack-scale systems. Over 10,000 NVIDIA employees are already using GPT-5.5-powered Codex internally. The model is rolling out to ChatGPT Plus, Pro, Business, and Enterprise — and to the API from April 24.
SpaceX struck a deal to co-develop a "coding and knowledge work AI" with Cursor, with an option to acquire the startup for $60 billion later this year — or pay $10 billion for the partnership work alone. The arrangement comes as xAI acknowledged that some SpaceX engineers have been slow to adopt Grok for coding because it lags rivals, with staffers quietly switching to Anthropic's Claude instead. Cursor co-founders Andrew Milich and Jason Ginsberg are joining xAI to rebuild Grok's coding capabilities from scratch.
Anthropic's newest flagship, Claude Opus 4.7, makes major strides in agentic coding and vision. Highlights include SWE-bench Verified 87.6% (up from 80.8%), GPQA Diamond 94.2% for scientific reasoning, and image support up to 3.75 megapixels — more than 3× the previous limit. A new xhigh effort level sits between high and max for finer reasoning-latency control, and task budgets let agents use a token ceiling gracefully. Pricing holds at $5/$25 per million tokens. The model is available on the Claude API, Amazon Bedrock, Google Vertex AI, and Microsoft Foundry.
At Google Cloud Next 2026, Google rebranded its Vertex AI developer environment as the Gemini Enterprise Agent Platform — a unified surface for building, deploying, governing, and optimizing AI agents. Key additions include a low-code Agent Studio, a code-first Agent Development Kit (ADK), cryptographic Agent Identity for each agent, and an Agent Gateway that enforces security policies against prompt injection and data leakage. Google also committed $750 million to its 120,000-member partner ecosystem to accelerate agentic AI adoption.
With Tim Cook's departure, incoming CEO John Ternus faces what analysts are calling Apple's most consequential strategic test: reversing the company's perceived AI lag. Apple is leaning on Google Gemini to power a reimagined Siri with on-screen awareness and cross-app integration, expected later in 2026. The company has deliberately avoided the massive data-center capex that rivals Microsoft, Google, Amazon, and Meta are committing — a bet on its edge AI and on-device approach that may define whether Apple Intelligence becomes a competitive differentiator or a liability.