Anthropic has agreed to spend $200 billion with Google Cloud over five years for compute and TPU access, making it the largest cloud commitment from a single AI company ever. The deal accounts for more than 40% of Google's disclosed revenue backlog and helped push Alphabet's market cap to $4.8 trillion. It cements the deeply intertwined financial relationship between the two companies, with Alphabet also investing up to $40 billion in Anthropic.
Apple is planning to let users swap in third-party AI models like Claude and Gemini to power Siri, Writing Tools, and Image Playground in iOS 27 this fall. The feature, internally called "Extensions," turns the iPhone into an AI marketplace where users pick their preferred provider through Settings. Google and Anthropic integrations are already being tested internally.
UK-based Google DeepMind employees have voted overwhelmingly to form what would be the first union at any frontier AI research lab. The move was triggered by Google's deal to let the Pentagon use Gemini models inside classified military networks for "any lawful purpose." Workers are demanding commitments against weapons development and stronger whistleblower protections.
Microsoft and OpenAI have formally moved to a non-exclusive licensing arrangement, allowing OpenAI to serve products on any cloud provider. Microsoft retains first-access rights through 2032 and the AGI clause has been dropped entirely. OpenAI has already begun striking multi-billion-dollar deals with Microsoft competitors including Amazon.
The Center for AI Standards and Innovation has signed agreements with Google DeepMind, Microsoft, and xAI to evaluate frontier models before public release, expanding on earlier partnerships with OpenAI and Anthropic. CAISI has already completed over 40 evaluations. The push was partly prompted by concerns over Anthropic's Mythos model and its cybersecurity implications.
A study published in Science by Harvard Medical School researchers found OpenAI's o1 model correctly diagnosed ER patients 67% of the time versus 50-55% for attending physicians, across 76 real cases from Beth Israel Deaconess Medical Center. The AI relied on text alone with no imaging or physical exam input. Researchers stopped short of recommending clinical deployment, calling for formal prospective trials.
Cloudflare is slashing 20% of its workforce even as it posted record quarterly revenue of $639.8 million, up 34% year-over-year. CEO Matthew Prince said AI efficiency gains eliminated the need for many support roles, and the company's internal AI usage surged 600% in three months. The cuts are part of a broader wave including Upwork (24% cut), Coinbase, and PayPal.
Alphabet briefly passed Nvidia by market cap this week and closed at $4.8 trillion, driven by Wall Street's conviction that Google owns "most of the stack" in AI — chips, models, cloud, and distribution. The Anthropic cloud deal, nearly doubled cloud backlog to $462 billion, and JPMorgan named it their top tech pick. All eyes now turn to Google I/O for the agent strategy reveal.
Nvidia's non-marketable equity securities ballooned to $22.25 billion from $3.39 billion a year earlier, with total commitments exceeding $40 billion. Major bets include a $30 billion stake in OpenAI, $3.2 billion in Corning, and $2.1 billion in data center operator IREN. Nvidia is rapidly becoming the AI industry's most active strategic investor on top of its hardware dominance.
IBM research shows the CAIO role has tripled in adoption in just one year, moving from a symbolic title to an operational C-suite position on par with CFO and COO. The role now centers on strategic alignment, reskilling workforces, and AI governance rather than running experimental pilots. CEOs expect the CAIO's influence to grow steadily through 2030.
