AI Radar: Daily Tech Brief

On this episode of AI RADAR, the industry faces a major fracture. **OpenAI** has lost over 1.5 million subscribers following a deal to deploy models for the Pentagon, a move made shortly after the U.S. government blacklisted **Anthropic** for refusing to compromise on safety guardrails.

Also in the news: **Meta** signs a massive $50 million annual licensing deal with News Corp to train Llama on content from the WSJ and NY Post, helping them avoid copyright litigation. Plus, **Perplexity** launches "Computer," an autonomous agent running 19 simultaneous models, and Google DeepMind reveals the high-speed **Nano Banana 2** image generator.

Sources:
• The Times of India: https://timesofindia.indiatimes.com/technology/tech-news/openai-loses-1-5-million-subscribers-in-less-than-48-hours-after-ceo-sam-altman-says-yes-to-the-deal-that-anthropic-rejected/articleshow/115891234.cms
• The Wall Street Journal: https://www.wsj.com/business/media/news-corp-meta-in-ai-content-licensing-deal-worth-up-to-50-million-a-year-54162460
• SitePoint: https://www.sitepoint.com/perplexity-computer-explained-safer-openclaw-ai-agents/
• Search Engine Land: https://searchengineland.com/google-nano-banana-2-promises-smarter-faster-image-generation-438123
• TechNode: https://technode.com/2026/03/02/deepseek-plans-v4-multimodal-model-release-this-week-sources-say/

Today on AI RADAR: **OpenAI** CEO Sam Altman admits the company's new Pentagon deal was "rushed" and is scrambling to amend terms to prohibit domestic surveillance, all while the administration officially blacklists rival **Anthropic**.

In hardware news, the chip wars escalate as **Meta** signs a historic $100 billion agreement with **AMD** to break free from **Nvidia's** dominance. Nvidia is countering immediately with a $4 billion investment in photonics to speed up data centers using light.

Plus: **Perplexity** launches its agentic "Computer" platform, and **Google DeepMind** unveils "Gemini Deep Think."

Sources:
• Axios: https://www.axios.com/2026/03/02/treasury-terminates-anthropic-ai-trump
• Marketplace: https://www.marketplace.org/2026/02/24/meta-100-billion-deal-buys-freedom-from-nvidia-dependence/
• Reuters: https://www.reuters.com/technology/nvidia-invest-4-billion-lumentum-coherent-2026-03-02/
• Business Insider: https://www.businessinsider.com/perplexity-ai-launches-computer-agentic-platform-2026-2
• Google DeepMind: https://deepmind.google/discover/blog/gemini-deep-think-redefining-scientific-research/

Today on AI RADAR: A high-stakes standoff escalates between **Anthropic** and the Pentagon as CEO Dario Amodei defies the Trump Administration's ban on the company's products. We break down the conflict over autonomous weapon safeguards and its impact on the AI defense sector.

Meanwhile, **OpenAI** shatters records with a massive $110 billion funding round led by Amazon, Nvidia, and SoftBank, catapulting its valuation to $730 billion. Plus, reports surface that **Nvidia** is preparing a "top secret" inference chip to counter custom silicon from Big Tech rivals.

Also: **Nvidia** forms a 6G AI alliance, and **Anthropic** accuses Chinese labs of model distillation.

Sources:
• TechPolicy.press: https://techpolicy.press/a-timeline-of-the-anthropic-pentagon-dispute/
• TheStreet: https://www.thestreet.com/technology/openai-just-raised-110-billion-and-it-still-may-not-be-enough
• SiliconANGLE: https://siliconangle.com/2026/03/01/report-nvidia-working-top-secret-ai-inference-chip-debut-next-month/
• Mobile World Live: https://www.mobileworldlive.com/nvidia-lures-operators-vendors-with-ai-6g-pitch
• TechInformed: https://techinformed.com/anthropic-accuses-chinese-ai-labs-of-claude-distillation/

The AI industry faces a major geopolitical shakeup as the Trump administration officially blacklists **Anthropic** over safety disputes, while **OpenAI** steps in to secure classified defense contracts with the newly renamed "Department of War."

In financial news, **Nvidia** smashes expectations with $68.1B in revenue, with CEO Jensen Huang declaring the arrival of the "Agentic AI" era. Meanwhile, **Google DeepMind** launches "Nano Banana 2," a high-speed image model designed for real-time creativity.

Plus: **Apple & Ericsson** partner on AI-driven 6G networks, and **Microsoft** unveils autonomous Copilot Tasks.

Sources:
• The Guardian: https://www.theguardian.com/technology/ai
• OpenAI News: https://openai.com/news
• Nvidia Reports: https://nvidianews.nvidia.com/
• Google Blog: https://blog.google/technology/ai/
• Ericsson: https://www.ericsson.com/en/press-releases
• Microsoft Blog: https://blogs.microsoft.com/

OpenAI secures a historic **$110 billion** funding round led by Nvidia and Amazon, rocketing its valuation to **$840 billion**. We discuss how this capital fuels the AI infrastructure race and cements OpenAI's dominance.

Meanwhile, President Trump orders US federal agencies to **cease using Anthropic** products. The directive follows a dispute over safety guardrails and national security, marking a major clash between AI ethics and defense requirements.

Also in this episode: Perplexity launches **"Computer"** to act as an AI operating system, Google brings memory features to free Gemini users, and Anthropic adjusts its safety scaling policy.

Sources:
• The Guardian: https://www.theguardian.com/technology/2026/feb/27/openai-announces-110bn-funding-round-valuation
• SiliconANGLE: https://siliconangle.com/2026/02/27/trump-orders-federal-agencies-stop-using-anthropic-products/
• TechCrunch: https://techcrunch.com/2026/02/27/perplexity-launches-computer-platform
• 9to5Google: https://9to5google.com/2026/02/26/gemini-app-past-chats-free-users/
• CBC News: https://www.cbc.ca/news/business/anthropic-amends-safety-policy-1.7126834

Today on AI RADAR: **Anthropic** takes a historic stand against the U.S. Pentagon, refusing a "final offer" to remove safety guardrails from Claude. CEO Dario Amodei risks a $200M contract to maintain ethical standards amidst national security pressures.

Meanwhile, **Nvidia** silences bubble fears with a massive $68.1B quarter. CEO Jensen Huang declares the industry has reached the "agentic AI inflection point," driven by autonomous systems rather than just chatbots.

Also in this episode: **Perplexity** pivots from search to action with its new "Perplexity Computer" platform, **Meta** signs a shock deal to rent Google's TPUs, and **DeepMind** releases the lightning-fast "Nano Banana 2" image model.

Sources:
• The Guardian: https://www.theguardian.com/technology/2026/feb/26/anthropic-pentagon-ai-safety-standoff
• RNZ News: https://www.rnz.co.nz/news/business/2026/02/26/nvidia-smashes-forecasts-with-record-quarter
• Elets CIO: https://cio.eletsonline.com/news/perplexity-ai-launches-perplexity-computer
• The Standard: https://www.thestandard.com.hk/breaking-news/section/4/20260226/meta-google-chip-deal
• Google Blog: https://blog.google/technology/ai/nano-banana-2-launch

Nvidia has silenced "AI bubble" fears with a massive **$68.1 billion quarter**, driven by what CEO Jensen Huang calls the new era of **"Agentic AI."** We break down the earnings and the insatiable demand for Blackwell Ultra chips.

Meanwhile, Google strikes back with **Gemini 3.1 Pro**, introducing a "Medium" thinking mode to balance speed and reasoning, plus new agentic capabilities for Android devices.

Also in this episode: OpenAI faces heavy scrutiny in Canada over the **Tumbler Ridge shooter case**, sparking debates on AI liability and a potential legal "duty to report" threats.

Plus: Microsoft pledges **$50 billion** for Global South infrastructure, and Anthropic partners with Pratham for education in India.

Sources:
• Nvidia Earnings: https://nvidianews.nvidia.com/news/nvidia-announces-financial-results-for-fourth-quarter-and-fiscal-2026
• Google Gemini 3.1: https://blog.google/products/gemini/gemini-3-1-pro-model-updates-android-features/
• OpenAI Canada Scrutiny: https://globalnews.ca/news/10293847/openai-tumbler-ridge-shooting-regulation-minister-solomon/
• Microsoft Investment: https://news.microsoft.com/2026/02/26/microsoft-announces-50-billion-investment-global-south-ai/
• Anthropic India: https://www.anthropic.com/news/pratham-partnership-india-education

In this episode of AI RADAR, we break down **Meta's** historic $100 billion deal with **AMD**, a strategic pivot designed to challenge Nvidia's monopoly on AI hardware.

We also discuss **Perplexity AI's** surprise decision to completely remove ads from its platform to prioritize user trust, and the **Pentagon's** ultimatum to **Anthropic** regarding the use of its models for military operations. Plus, **Samsung** integrates Perplexity into the Galaxy S26 and **Google Vids** gets a multilingual upgrade.

Sources:
• Meta/AMD Deal: https://www.latimes.com/business/story/2026-02-24/meta-to-buy-ai-chips-from-amd-in-100-billion-deal
• Perplexity Removes Ads: https://www.tomsguide.com/ai/perplexity-just-removed-ads-to-protect-trust-heres-why-chatgpt-should-do-the-same
• Pentagon vs Anthropic: https://www.piquenewsmagazine.com/world-news/hegseth-warns-anthropic-to-let-the-military-use-the-companys-ai-tech-as-it-sees-fit-ap-sources-say
• Samsung Galaxy S26: https://www.deccanherald.com/technology/technology-news/samsungs-android-flagship-galaxy-s26-series-to-support-perplexity-ai
• Google Vids Update: https://workspaceupdates.googleblog.com/2026/02/ai-avatars-and-ai-voiceovers-in-google-vids-now-available-in-seven-new-languages.html

On this episode of AI RADAR, tensions escalate globally. **Anthropic** formally accuses major Chinese labs of stealing Claude's reasoning capabilities through "industrial-scale" model distillation, labeling it a national security risk.

Meanwhile, **OpenAI** faces political fury in Canada after revelations it banned a school shooter's account months before the attack but didn't alert authorities.

Plus, the **Pentagon** issues an ultimatum to Anthropic over autonomous weapons restrictions, threatening to label the startup a "supply chain risk" while turning to Elon Musk’s **xAI**.

Also: **Meta** revives its AI smartwatch plans and **Google DeepMind** warns of a critical hardware choke point.

Sources:
• Anthropic Blog: https://www.anthropic.com/news/detecting-and-preventing-distillation-attacks
• CBC News: https://www.cbc.ca/news/canada/british-columbia/eby-tumbler-ridge-shooting-openai
• Axios: https://www.axios.com/technology/artificial-intelligence
• The Information: https://www.theinformation.com/topics/meta
• CNBC: https://www.cnbc.com/artificial-intelligence

Today on AI RADAR: **Anthropic** disrupts the cybersecurity market with "Claude Code Security," a new agent that reasons through code logic to find vulnerabilities. The release triggered a sell-off for major stocks like CrowdStrike and Cloudflare.

Also, **Samsung** makes a bold move by integrating **Perplexity** as a native system-level agent on Galaxy devices, ending its exclusive reliance on Google. Meanwhile, **Nvidia** re-enters the PC market with AI chips to challenge Apple and expands its infrastructure partnership with **Meta**.

Plus, **Google** launches Gemini 3.1 Pro with enhanced reasoning, and **OpenAI** targets a staggering $600 billion in compute spending by 2030.

Sources:
• Security Affairs: https://securityaffairs.com/anthropic-unveils-claude-code-security-detect-fix-code-bugs/
• The Tech Buzz: https://thetechbuzz.com/samsung-opens-galaxy-ai-perplexity-multi-agent-push/
• Seeking Alpha: https://seekingalpha.com/news/nvidia-re-enters-pc-market-ai-powered-laptop-chips
• Google The Keyword: https://blog.google/technology/ai/gemini-3-1-pro-smarter-model/
• Reuters: https://www.reuters.com/technology/artificial-intelligence/openai-aims-600-billion-computing-power-spending-2030