TL;DR
OpenAI raises $8.3B at $300B valuation; Google launches Gemini 2.5 Deep Think; Amazon, Apple, Alphabet boost AI capex.
Highlights
-
- Google launches Gemini 2.5 Deep Think, a multi-agent reasoning model for $250/month Ultra subscribers, outperforming recent models; API access for developers coming soon 2.
-
- Amazon lifts 2025 capex outlook to $118B, with Q2 spend at $31.4B—mostly for AWS AI infrastructure and custom chips; hyperscaler capex projected above $345B this year 3.
-
- Anthropic overtakes OpenAI in enterprise LLM API usage (32% vs. 25%) as enterprise LLM spend doubles to $8.4B in H1 2025; FAL raises $125M to expand generative infrastructure 7.
-
- U.S. Army consolidates 75 software contracts in a $10B, 10-year Palantir deal for data analytics and AI tools, streamlining procurement 5.
-
- China’s NDRC pledges further support for AI commercialization and integration with traditional industries, alongside broader economic stimulus 9.
-
- Google loses Epic antitrust appeal; must allow rival app stores and payment systems on Android within eight months 10.
-
- Microsoft study finds interpreters, historians, and analysts most at risk from AI, while manual labor and specialized healthcare roles remain secure 11.
-
- Figma surges over 200% in NYSE debut, reflecting renewed investor interest in AI-adjacent SaaS IPOs; generative AI not seen as immediate threat to core tools 12.
-
- Reddit posts first GAAP profit; AI-powered search and data licensing to AI developers drive revenue growth 13.
Commentary
The latest funding and earnings data confirm that capital deployment into AI infrastructure and products remains at historic highs. OpenAI’s $8.3B raise at a $300B valuation 1, combined with hyperscaler capex projections exceeding $345B for 2025 3, signals continued prioritization of compute build-out and model scaling. Amazon , Alphabet , and Apple all posted strong quarters, with each company highlighting expanding AI infrastructure as a strategic focus 814. However, investor scrutiny is rising on how quickly these investments will translate into earnings, as seen in Amazon ’s post-earnings share decline despite robust cloud revenue 15.
On the product and adoption front, Google ’s Gemini 2.5 Deep Think brings multi-agent reasoning to high-end users, setting a new performance bar but also underscoring the high compute costs that limit access to premium tiers 2. Anthropic’s overtaking of OpenAI in enterprise LLM API usage, especially in code generation, reflects a rapidly shifting competitive landscape where enterprise buyers are willing to switch providers for performance and reliability 7. The dominance of proprietary models persists, with open-source adoption declining among large-scale enterprise workloads 7.
Regulatory and policy developments are accelerating. OpenAI’s removal of ChatGPT’s public sharing feature highlights ongoing privacy and data security concerns 6, while Google ’s loss in the Epic antitrust case will force major changes to Android app distribution and payments 10. In parallel, China’s NDRC is emphasizing AI commercialization and rational competition as part of broader economic support 9, and U.S. government procurement is consolidating around large-scale, flexible contracts, as seen in the Army’s $10B Palantir deal 5.
The IPO market is showing renewed interest in AI-adjacent SaaS, as Figma’s 200%+ debut demonstrates 12, while Reddit’s first profit underscores the growing value of AI-powered search and data licensing 13. Microsoft ’s occupational risk study provides further clarity on which job categories are most exposed to AI-driven disruption 11.