AI Radar: Daily Tech Brief

In this episode of AI RADAR, we break down the most critical developments in the artificial intelligence race.

**Anthropic** has announced a major policy to cover 100% of grid upgrade costs for its data centers, ensuring residential consumers aren't hit with the bill. Meanwhile, **OpenAI** has officially permitted the U.S. Pentagon access to ChatGPT for "all lawful uses," a significant shift from previous restrictions.

Internally, reports describe a **"Code Red"** atmosphere at OpenAI, with 24/7 shifts implemented to counter the rise of Google's Gemini 3 and Anthropic. We also touch on **Nvidia's** strict export guardrails for China and OpenAI's latest engineering insights.

Sources:
• The Register: https://www.theregister.com
• Semafor: https://www.semafor.com
• TechCrunch: https://techcrunch.com
• Reuters: https://www.reuters.com
• OpenAI Blog: https://openai.com/news

Today on AI RADAR: Video AI pioneer Runway secures a massive $315M Series E led by General Atlantic, **Nvidia**, and **AMD**, doubling its valuation to **$5.3 billion** as it pivots to building "World Models" for physics simulation.

Also: Anthropic achieves a stunning milestone with **Claude Opus 4.6**, deploying a swarm of autonomous agents to build a functional C compiler from scratch—capable of compiling the Linux kernel—for just $20,000 in API credits.

Plus: The Indian government enforces a strict **two-hour** deadline for platforms like Meta and Google to remove deepfakes, threatening legal immunity for non-compliance.

In other news: Discord faces backlash over AI age verification, and OpenAI's custom chip roadmap with Broadcom hits delays.

Sources:
• SiliconANGLE: https://siliconangle.com/2026/...
• Ars Technica: https://arstechnica.com
• TechCrunch: https://techcrunch.com
• Tom's Hardware: https://www.tomshardware.com

Today on AI RADAR: OpenAI aggressively targets the enterprise with **Frontier**, a new platform for autonomous agents, while confirming that **ads are coming to ChatGPT**. Meanwhile, the race for AGI intensifies as Anthropic nears a massive **$20B raise** ($300B valuation), though the milestone is overshadowed by a top safety researcher resigning with a warning that the **"world is in peril."**

Also in this episode:
• The EU threatens **antitrust action** against Meta for restricting rival AIs on WhatsApp.
• Google releases the updated **Gemini 2.5 Flash** model.
• Nvidia CEO Jensen Huang calls fears of AI replacing software stocks **"illogical."**

Sources:
• TechCrunch: https://techcrunch.com
• Forbes: https://www.forbes.com
• The Guardian: https://www.theguardian.com
• Google Blog: https://blog.google
• TheStreet: https://www.thestreet.com

On this episode of AI RADAR, we analyze the deepening talent drain at **OpenAI**. Following Sam Altman's "Code Red" directive, senior researchers are departing as the company pivots resources from experimental projects to ChatGPT product speed.

Also: **Nvidia** files a motion to dismiss a major copyright lawsuit, denying claims that it knowingly used "Anna's Archive" to train its models. Meanwhile, **Waymo** officially integrates DeepMind's **Genie 3** to simulate rare driving scenarios—like tornadoes or highway landings—using generative world models.

Plus: **ByteDance** releases Protenix-v1 to rival AlphaFold, and **Goldman Sachs** partners with **Anthropic**.

Sources:
• Tom's Hardware: https://www.tomshardware.com/tech-industry/artificial-intelligence/nvidia-says-it-didnt-use-pirated-books-to-train-its-ai-models-company-asking-for-annas-archive-suit-to-be-dismissed
• Times of India: https://timesofindia.indiatimes.com/technology/tech-news/googles-waymo-on-how-deepminds-ai-model-is-preparing-its-robotaxis-for-difficult-situations-it-allows-us-to-explore/articleshow/118000000.cms
• Computing UK: https://www.computing.co.uk/news/4333333/openais-focus-chatgpt-prompts-senior-staff-departures
• Reddit/AI News: https://www.reddit.com/r/ArtificialInteligence/comments/1io4444/oneminute_daily_ai_news_282026/
• Bush AI Cave: https://bushaicave.com/2026/02/08/one-minute-daily-ai-news-2-8-2026/

Is the AI bubble real? In this episode of AI RADAR, **Nvidia CEO Jensen Huang** pushes back against skeptics, calling the massive capital expenditure in tech "sustainable" and triggering an 8% stock rally.

We also cover **China's regulatory crackdown** on counterfeit DeepSeek and ChatGPT services, and **Apple's strategic decision** to cancel its standalone AI Health Coach project. Plus, **Google Gemini** gets a UI refresh and leaks reveal **OpenAI** took a $5 billion loss in 2024.

Sources:
• Tech in Asia: https://www.techinasia.com/news/nvidia-ceo-ai-infrastructure-spending-remains-sustainable
• South China Morning Post: https://www.scmp.com/tech/policy/article/3336712/china-cracks-down-fake-chatgpt-deepseek-ai-services-latest-malpractice-probe
• Engadget: https://www.engadget.com/general/apple-reportedly-axed-its-planned-ai-health-coach-133038627.html
• 9to5Google: https://9to5google.com/2026/02/07/google-gemini-tools-my-stuff-redesign/
• CNBC: https://www.cnbc.com/2026/02/07/openai-5-billion-loss-revenue-growth.html

Today on AI RADAR: The software market is bleeding as **Anthropic** launches **Claude Opus 4.6** and "Cowork" agents, wiping **$300 billion** off legacy SaaS valuations. Investors are betting that agentic AI replaces the seat-based subscription model.

Meanwhile, **OpenAI** strikes back with its "Frontier" platform and **GPT-5.3-Codex** to solve enterprise governance, while **Google DeepMind** claims a "Holy Grail" breakthrough with "Reinforcement Attention Learning" to fix AI memory retention.

Plus: **Perplexity** ditches AWS for a $750M deal with Microsoft, and **Nvidia** surges as hardware demand accelerates.

Sources:
• Anthropic Newsroom: https://www.anthropic.com/news/claude-opus-4-6
• OpenAI Announcements: https://openai.com/index/introducing-gpt-5-3-codex/
• Google DeepMind Blog: https://deepmind.google/discover/blog/
• Bloomberg Technology: https://www.bloomberg.com/technology
• Nvidia Newsroom: https://nvidianews.nvidia.com/

Is the $100 billion alliance between **Nvidia** and **OpenAI** falling apart? Reports suggest the deal is on the rocks due to hardware disputes, sending Nvidia stock tumbling. We discuss the implications for the AI supply chain.

Also in this episode:

• **Google Gemini** surpasses 750 million monthly users as the **Gemini 3** model integrates deeper into Android and Chrome.
• **OpenAI** faces a "brain drain" of senior researchers amid a cultural shift from AGI research to consumer product stability.
• **ChatGPT** suffers recurring outages due to fine-tuning errors.
• The US blocks Nvidia H200 chip exports to China again over security concerns.

Sources:
• The Guardian: https://www.theguardian.com/te...
• The Guardian: https://www.theguardian.com/te...
• Times of India: https://timesofindia.indiatime...
• Fox West Texas: https://www.foxwesttexas.com/n...
• Asia Financial: https://www.asiafinancial.com/...

Google's AI dominance is growing. On this episode of AI RADAR, we break down Alphabet's Q4 earnings where Sundar Pichai revealed **Gemini** now has **750 million monthly active users**.

Also, **Anthropic** takes a hard stance: **Claude** will remain ad-free forever, calling ads "incompatible" with helpful AI. Meanwhile, **Perplexity** shakes up the cloud market with a **$750M deal with Microsoft**, shifting some workload to Azure.

Finally, we touch on **ChatGPT**'s recent outages and Google DeepMind's new world-building AI, **Project Genie**.

Sources:
• The Guardian: https://www.theguardian.com/te...
• Anthropic Newsroom: https://www.anthropic.com/news
• Binance News: https://www.binance.com/en/squ...
• TechRadar: https://www.techradar.com/comp...
• Google Blog: https://blog.google/technology...

Is the most dominant alliance in AI fracturing? Reports indicate **OpenAI** is clashing with **Nvidia** over inference costs, stalling a potential $100B investment. Meanwhile, the "AI Cold War" escalates as US lawmakers introduce legislation to ban the Chinese model **DeepSeek** from government devices.

Plus, **Anthropic** makes a massive play for the developer ecosystem by integrating the Claude Agent SDK directly into Apple's Xcode, and rumors swirl about an imminent "Sonnet 5" release.

**Also in this episode:**
• Mozilla Firefox adds a "GenAI Kill Switch" for privacy
• Claude Sonnet 5 leaks via API glitches

Sources:
• OpenAI & Nvidia Strain: https://techcrunch.com/openai-...
• US Bans DeepSeek: https://reuters.com/us-lawmake...
• Anthropic & Apple Xcode: https://anthropic.com/news/cla...

On this episode of AI RADAR:

**humans& Launches with $480M**: A new AI lab, founded by former OpenAI and DeepMind researchers, has secured a massive seed round led by 500 Global. Unlike competitors, they are focusing on "human-centric" models rather than raw AGI scaling.

**RIP GPT-4o**: OpenAI announces the permanent retirement of GPT-4o by Feb 13, 2026. Users are being forced onto GPT-5.2, effectively deleting the "sycophantic" personality that 0.1% of users refused to leave.

**Nvidia Investment Stalls**: Nvidia shares dip as a planned $100B investment in OpenAI hits a roadblock over hardware disagreements, with OpenAI exploring alternatives like Cerebras.

Plus: **Oracle** seeks $50B for cloud capacity, and **SoftBank & Intel** launch "Z-Angle Memory" to lower power consumption.

Sources:
• humans& Launch: https://500.co
• OpenAI Updates: https://openai.com
• Nvidia Report: https://theinformation.com
• Oracle News: https://oracle.com
• Intel News: https://intel.com