AI Weekly - Week 02/2026
TL;DR
This week in 30 seconds:
- Health AI: OpenAI launches ChatGPT Health with access to medical data – 230 million users already ask health questions weekly
- Hardware Push: Nvidia unveils Rubin Platform at CES with 10x cheaper inference costs
- Enterprise Drive: Anthropic secures Allianz as partner and targets $350 billion valuation
- AI Everywhere: Google brings Gemini to Gmail, Microsoft to retail, while CES shows where we’re heading
Audio Version
📑 Chapters
Read aloud with edge-tts (Microsoft Azure Neural Voice: en-US-AndrewNeural)
🌟 Story of the Week
Image: AI-generated with Pollinations.ai
OpenAI Brings AI to Healthcare
On January 7, 2026, OpenAI announced ChatGPT Health – a dedicated AI environment for health questions with medical data integration [1]. The numbers show the demand: Over 230 million users already ask health questions on the platform weekly [1].
The new feature allows connection with Apple Health, MyFitnessPal, Function Labs, and other health apps through partner b.well [1]. Conversations remain in an isolated environment and are not used for model training [1].
“ChatGPT Health addresses healthcare system challenges including cost and access barriers, overbooked doctors, and a lack of continuity in care.” [1]
— Fidji Simo, CEO of Applications at OpenAI
Critical Voices: OpenAI’s own Terms of Service emphasize that the tool is “not intended for diagnosis or treatment” [1]. Large Language Models are prone to hallucinations – potentially dangerous in medicine. Regulatory questions remain open: No FDA approval was mentioned, nor liability issues for erroneous AI recommendations. Privacy advocates will likely question long-term storage of health data – despite HIPAA compliance for the Enterprise version.
In parallel, OpenAI announced “OpenAI for Healthcare” for organizations – with HIPAA compliance and special enterprise features. Rollout begins in the coming weeks [1].
Bottom Line: OpenAI makes the first major move into the healthcare market, but regulatory and ethical hurdles are enormous. Anyone making medical decisions here should proceed with caution.
🔥 More Top Stories
1. Nvidia Shows Future of AI Hardware at CES
Image: AI-generated with Pollinations.ai
Nvidia unveiled the Rubin Platform on January 5 – six integrated chips for the next generation of AI supercomputers [2]. The system combines the Vera CPU with 88 Olympus cores, the Rubin GPU with 50 petaflops of NVFP4 compute, and four additional components into an integrated superchip [2].
The performance leaps are remarkable: 10x lower inference token costs and 4x fewer GPUs for training Mixture-of-Experts models compared to Blackwell [2]. Rubin is already in production, systems arriving in the second half of 2026 [2].
Major cloud providers are on board: AWS, Google Cloud, Microsoft, Oracle, as well as Anthropic and OpenAI as first customers [2].
2. Google Turns Gmail into AI Hub
Image: AI-generated with Pollinations.ai
On January 8, Google announced Gemini 3 integration in Gmail [3]. New features include AI Overviews for thread summaries, “Help Me Write” for composing emails, and context-based Smart Replies [3].
Particularly interesting: Gmail can now answer questions like “Who was the plumber who gave me a quote last year?” by searching the entire email history [3]. Thread summaries are free for everyone, advanced features require Google AI Pro or Ultra subscriptions [3].
In parallel, Google showed new Gemini features for Google TV at CES – with voice control for settings and Google Photos integration [10].
3. Anthropic on Enterprise Mission
Image: AI-generated with Pollinations.ai
Anthropic kicked off 2026 with a bang: On January 9, the partnership with insurance giant Allianz was announced [4]. Claude Code will be rolled out to all Allianz employees, plus both companies are jointly developing custom AI agents for insurance workflows [4].
According to a December Menlo Ventures survey, Anthropic already holds 40% enterprise market share – up from 32% in July 2025 [4].
In parallel, reports circulate about a $10 billion funding round that would value Anthropic at $350 billion [8]. Lead investors reportedly include Singapore’s sovereign wealth fund GIC and Coatue Management, with Microsoft and Nvidia as existing investors [8].
⚡ Quick Hits
Brief notes:
- Meta Nuclear: Meta secured 6+ gigawatts of nuclear power from Oklo, TerraPower, and Vistra to supply AI infrastructure [5]
- Grok Ban: Indonesia blocked xAI’s Grok due to non-consensual sexualized deepfakes [6]
- Microsoft Retail: Microsoft announced Copilot Checkout – AI shopping with PayPal, Shopify, and Stripe integration, available in the US [7]
- Audio-First: OpenAI preparing an audio-first personal device, expected in about a year [9]
- OpenAI Grove: New accelerator program for early-stage founders runs from January 22 to February 27, 2026
🛠 Tool of the Week
Claude Code v2.1.0 - AI Coding Agent with Improved Agent Lifecycle Control
Anthropic’s coding agent received a major update with 1,096 commits [11]. New features: Better session portability, skills development, multilingual outputs, and refined agent management [11]. According to VentureBeat, developers say Claude Code outperforms Cursor, GitHub Copilot, and even Gemini 3 Pro for coding.
💥 Fail of the Week
“Grok Generates Non-Consensual Deepfakes”
Indonesian authorities temporarily blocked xAI’s Grok chatbot after it became known that the platform can create sexualized deepfakes without consent of depicted persons [6].
Root Cause: Missing content moderation and safety guardrails in image generation – a problem affecting many AI image generators, but especially critical for a public chatbot.
What We Learn: Safety cannot be retrofitted – guardrails must be part of the architecture from day one.
📊 Number of the Week
$350 Billion
The target valuation for Anthropic in the current funding round [8]. For comparison: That would exceed Intel’s market cap ($200B) and make Anthropic one of the most valuable private tech companies. The $10 billion investment would primarily flow into compute infrastructure and model training.
📚 Reading List
For the weekend:
- In 2026, AI will move from hype to pragmatism - TechCrunch on the shift to smaller, task-focused models (8 min)
- NVIDIA CES 2026 Special Presentation - Jensen Huang’s complete Rubin Platform keynote (15 min)
- The creator of Claude Code just revealed his workflow - Inside look at Claude Code development (10 min)
🔮 Next Week
What’s coming:
- OpenAI Grove Accelerator starts January 22
- Rubin-based cloud instances expected for Q3/Q4 2026
- ChatGPT Health rollout in coming weeks
- More CES aftershocks: Hands-on reviews of announced AI hardware
🤖 Behind This Newsletter
Generated in: ~40 minutes Sources scanned: 47 articles from 12 feeds Stories found: 20 → 7 selected Validation: 4 Agents, 3 corrections Model: Claude Opus 4.5 + Haiku (Validation) Images: Pollinations.ai (5 generated)
Full Metrics
| Phase | Metric | Value |
|---|---|---|
| Source Collection | RSS Feeds | 12 |
| Source Collection | WebSearch Queries | 5 |
| Selection | Stories Presented | 20 |
| Selection | Stories Selected | 7 |
| Draft | Words | ~1900 |
| Draft | Sources Cited | 11 |
| Validation | Fact-Check Issues | 1 (minor) |
| Validation | Balance Issues | 8 |
| Validation | Quality Issues | 3 |
| Validation | Legal Issues | 0 |
This newsletter was researched and written with AI assistance. Images generated with Pollinations.ai.
Sources
- OpenAI unveils ChatGPT Health, says 230 million users ask about health each week
- NVIDIA Kicks Off the Next Generation of AI With Rubin
- Gmail launches AI features like AI Overviews and more, made possible by Gemini 3
- Anthropic adds Allianz to growing list of enterprise wins
- Meta signs deals with three nuclear companies for 6 plus gw of power
- Indonesia blocks Grok over non-consensual sexualized deepfakes
- Microsoft propels retail forward with agentic AI capabilities
- Claude maker Anthropic to raise $10bn round at $350bn valuation
- OpenAI bets big on audio as Silicon Valley declares war on screens
- Google previews new Gemini features for TV at CES 2026
- Anthropic's Claude Code 2.1.0 arrives with smoother workflows and smarter agents