Today’s AI news on April 29, 2026 spans frontier model releases, unprecedented corporate acquisitions, and a historic space milestone. Google released Gemma 4 under Apache 2.0, OpenAI acquired its first media property, Novo Nordisk announced a full-company AI partnership with OpenAI, and NASA confirmed the first Mars drives ever planned by artificial intelligence.
Google Releases Gemma 4 Open-Source Model Under Apache 2.0
Google launched Gemma 4, its latest series of open models built specifically for advanced reasoning and agentic workflows. Released under the Apache 2.0 license, Gemma 4 is immediately available for commercial deployment without licensing fees. The model is designed for multi-step reasoning tasks and agent orchestration scenarios — the same capabilities Google is commercializing via its Gemini Enterprise Agent Platform. Open-sourcing a version of these capabilities signals Google’s intent to build developer ecosystem lock-in while monetizing the enterprise layer.
OpenAI Acquires TBPN: Its First Media Property
OpenAI acquired TBPN — Technology Business Programming Network — a daily Silicon Valley talk show on track to generate $30 million in revenue this year. This is OpenAI’s first acquisition of a media property and marks a strategic pivot toward owning content distribution channels. The move comes as OpenAI faces scrutiny over its missed 2026 revenue targets (reported by WSJ on April 28), suggesting the company is diversifying its revenue base beyond API subscriptions and ChatGPT Plus.
Novo Nordisk Partners with OpenAI for Full-Company AI Deployment
Danish pharmaceutical giant Novo Nordisk announced a strategic partnership with OpenAI to integrate AI across its entire business — from drug discovery and clinical trials to manufacturing, supply chains, and commercial operations — with full deployment planned by end of 2026. This is one of the most ambitious pharma-AI partnerships announced to date. Novo Nordisk joins a growing list of Fortune 500 companies deploying OpenAI models at the enterprise level, even as OpenAI’s consumer growth comes under question.
Microsoft Commits $10 Billion to Japan AI Infrastructure
Microsoft announced a four-year, $10 billion investment in Japan (2026–2029) covering AI data center expansion in partnership with SoftBank and Sakura Internet, deep cybersecurity cooperation with the Japanese government, and a pledge to train over one million Japanese engineers and developers by 2030. The announcement underscores that global AI infrastructure buildout continues at full speed despite OpenAI’s revenue challenges — the infrastructure layer is multi-year committed spending, not discretionary.
NASA Mars: First AI-Planned Rover Drives Completed
NASA confirmed that its Perseverance rover completed the first Mars drives ever planned by artificial intelligence. The system used Anthropic’s Claude vision-language models to analyze orbital imagery and terrain data and autonomously generate safe waypoints. The rover successfully navigated to targets selected entirely by AI — a milestone that demonstrates real-world autonomy in the most demanding environment possible. This builds on earlier AI science instrument scheduling and represents a significant step toward fully autonomous planetary exploration.
Stanford AI Index 2026: Generative AI Reaches 53% Population Adoption
The Stanford HAI AI Index 2026 report released this week reveals generative AI reached 53% population adoption within three years — faster than the personal computer or the internet. The estimated value of generative AI tools to U.S. consumers reached $172 billion annually by early 2026. The report also found that AI performance on hard reasoning benchmarks improved more in 2025 than in the previous five years combined, suggesting that capability scaling has not plateaued despite concerns about training data exhaustion.
DeepSeek Cuts API Prices 90% Again
DeepSeek permanently reduced input cache prices across its entire API series to one-tenth of previous costs this week, continuing its aggressive pricing strategy. Total LLM inference costs have now dropped roughly 50% versus January 2026 across the market. For enterprises evaluating LLM providers, the cost landscape is shifting faster than annual budget cycles can accommodate — organizations that locked in multi-year API contracts at 2025 prices are paying a significant premium.
