AI-generated illustration for Week 02/2026 focusing on Health AI and Enterprise Adoption
Image generated with Pollinations.ai
Weekly Briefing 11 min read

AI Weekly - Week 02/2026

Sunday, January 11, 2026

This article was researched and written with AI

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

0:00 0:00
Speed:
Download MP3
📑 Chapters

Read aloud with edge-tts (Microsoft Azure Neural Voice: en-US-AndrewNeural)


🌟 Story of the Week

OpenAI ChatGPT Health 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

Nvidia Rubin Platform 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

Google Gemini Gmail 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

Anthropic Allianz Partnership 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.

Claude Code on GitHub


💥 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:

  1. In 2026, AI will move from hype to pragmatism - TechCrunch on the shift to smaller, task-focused models (8 min)
  2. NVIDIA CES 2026 Special Presentation - Jensen Huang’s complete Rubin Platform keynote (15 min)
  3. 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
PhaseMetricValue
Source CollectionRSS Feeds12
Source CollectionWebSearch Queries5
SelectionStories Presented20
SelectionStories Selected7
DraftWords~1900
DraftSources Cited11
ValidationFact-Check Issues1 (minor)
ValidationBalance Issues8
ValidationQuality Issues3
ValidationLegal Issues0

This newsletter was researched and written with AI assistance. Images generated with Pollinations.ai.