This Week in AI & Automation

Week of March 10, 2026

Anthropic sued the Pentagon after being blacklisted for refusing to let Claude power autonomous weapons. Atlassian cut 1,600 employees and replaced its CTO with two AI-focused executives. Washington state passed the first chatbot safety laws in the US. And Jensen Huang is about to take the GTC stage on Monday with what he calls "a chip to surprise the world." This week's theme: the fight over who controls AI — and what it's allowed to do — got real.

The Big Story

Anthropic Sues the Pentagon Over "Supply Chain Risk" Blacklisting

On March 9, Anthropic filed two federal lawsuits against the Trump administration after the Pentagon designated it a "supply chain risk" — a label normally reserved for Chinese telecom companies and foreign adversary contractors. The reason: CEO Dario Amodei refused to allow Claude to be used for autonomous weapons systems or mass surveillance.

The designation effectively bars federal agencies from purchasing Anthropic products and signals to defense contractors that working with the company carries regulatory risk. Within 24 hours, over 30 employees from OpenAI and Google DeepMind signed a public statement supporting Anthropic's position — a rare act of cross-company solidarity in the AI industry.

Google and Microsoft both confirmed they will continue working with Anthropic on non-defense projects. But the message from the Pentagon is clear: if you build AI and refuse military applications, there are consequences.

Source: NPR, TechCrunch, CNBC

Our Take: This is the most consequential AI-government conflict yet — and it matters for every enterprise buyer. If you're evaluating AI vendors, the Anthropic situation forces a question you couldn't ignore anymore: does your AI vendor's safety stance create business risk for you? For companies building on Claude, the lawsuit creates short-term uncertainty in government-adjacent sectors. But it also proves something: Anthropic is willing to lose revenue to maintain safety commitments. In a market full of vendors who say anything to close a deal, that's worth knowing.

Notable Developments

Atlassian Cuts 1,600 Jobs — Replaces CTO with Two AI Executives

Atlassian announced on March 11 that it is laying off 1,600 employees — 10% of its global workforce — to "self-fund" AI investments. But the layoffs aren't the real story. The CTO restructuring is.

CTO Rajeev Rajan is stepping down, replaced by two AI-focused CTOs: Taroon Mandhana (CTO Teamwork) and Vikram Rao (CTO Enterprise). Atlassian is literally splitting its technical leadership along an AI axis. The restructuring will cost $225-236 million in charges.

This follows Block (Square) doing essentially the same thing last month — cutting staff and reorganizing leadership around AI capabilities. A pattern is forming: enterprise software companies aren't adding AI to their products. They're rebuilding their organizations around AI as the core operating model.

Source: Bloomberg, CNBC

Our Take: If your company uses Jira, Confluence, or Bitbucket, expect significant product changes in the next 12 months. Atlassian isn't doing incremental AI features — they're restructuring the entire company around it. For enterprises planning AI transformation, this is the pattern to watch: AI isn't a product feature you add. It's an organizational shift that changes roles, leadership structure, and headcount allocation.

Washington State Passes First US Chatbot Safety Laws

Washington state gave final approval to two bills on March 12-13 that establish the first state-level chatbot safety requirements in the US. HB 1170 requires AI-generated images, audio, and video to include embedded watermarks and mandates detection tools for platforms with over 1 million users. HB 2225 requires hourly reminders that users are talking to AI (not humans), suicide ideation detection protocols, and prohibits chatbots from showing explicit content to minors.

Both bills head to Governor Ferguson's desk. Meanwhile, the federal government hit two regulatory deadlines on March 11: the Secretary of Commerce must publish evaluations of state AI laws that conflict with federal policy, and the FTC must issue guidance on how the FTC Act applies to AI.

Source: Transparency Coalition, Mondaq

Our Take: Companies deploying AI for customer support or conversational commerce need to pay attention. The hourly disclosure requirement — reminding users they're talking to AI — could become a template for other states. If you're building AI voice agents, design for disclosure from day one. Retrofitting compliance is always more expensive than building it in.

Quick Hits

• Apple's Gemini-powered Siri ships in iOS 26.4: The rebuilt Siri uses Google's 1.2 trillion parameter Gemini model with on-screen context awareness, multi-step action chaining, and natural conversations — all running on Apple's Private Cloud Compute servers. The largest AI assistant deployment by device count is about to get dramatically smarter. 9to5Mac
• ChatGPT hits 900M weekly active users: a16z's latest Top 100 Gen AI Apps report shows ChatGPT doubled its user base in one year. Claude's paid subscriber growth is 200%+ YoY. AI notetakers (Fireflies, Fathom, Otter, Granola) have hit 20M combined visitors. a16z
• Global VC funding hit $189B in February: The largest single month in venture history, with 780% YoY increase. Over 40% of seed and Series A funding went to rounds of $100M or more. The capital bar for AI startups keeps rising. Crunchbase
• 88% of enterprises now use AI in at least one function: Deloitte's latest survey confirms AI adoption has crossed mainstream thresholds. The question is no longer "are you using AI" — it's "how many functions have you automated."

Numbers of the Week

What We're Watching

NVIDIA GTC kicks off Monday. Jensen Huang delivers his keynote March 16 at 11am PT in San Jose. Pre-conference reporting points to the Vera Rubin GPU architecture (successor to Blackwell), a CPU-only rack designed specifically for agentic AI inference, and NemoClaw — Nvidia's platform for deploying AI agents across enterprise systems. The CPU pivot is the most interesting signal: Nvidia is acknowledging that not all AI workloads need GPUs, particularly inference for production AI agents. Whatever Huang unveils sets the AI infrastructure playbook for the next 18 months.

The Anthropic-Pentagon conflict will shape enterprise AI procurement for years. If the courts side with the Pentagon, every AI company faces a choice: comply with military use cases or risk losing government market access. If Anthropic wins, it establishes that AI companies can maintain safety guardrails without regulatory punishment. Either outcome changes how enterprises evaluate AI vendor risk.

AI-driven workforce restructuring is accelerating. Atlassian and Block aren't outliers — they're early movers. When enterprise software companies with tens of thousands of employees restructure their entire leadership around AI, it signals that the "AI as a feature" era is ending. The next phase is "AI as the org chart." Every company running AI transformation should be asking: what does our leadership structure look like in 12 months?

The Bottom Line

This was the week AI stopped being a technology conversation and became a power struggle. Anthropic drew a line on safety and the Pentagon pushed back with the heaviest regulatory weapon available. Atlassian decided AI was important enough to fire 1,600 people and restructure its entire technical leadership. Washington state decided AI chatbots need safety laws before someone gets hurt. And next week at GTC, Huang will reveal the hardware that makes the next generation of all this possible.

The common thread: institutions are grappling with the fact that AI is no longer optional, controllable, or ignorable. Governments want to regulate it. Companies are reorganizing around it. And the AI companies themselves are fighting over what it's allowed to do. For enterprises still treating AI as a department-level initiative, the message is clear: this is a board-level decision now.

Get This Week in AI & Automation delivered every Saturday.

Previous Editions

• This Week in AI & Automation: The Computer Use Era (Mar 7, 2026)
• This Week in AI & Automation: The 40% Cut (Feb 28, 2026)
• This Week in AI & Automation: The $100 Billion Week (Feb 21, 2026)
• This Week in AI & Automation: The Age of Agent Teams (Feb 14, 2026)
• This Week in AI & Automation: The $470 Billion Question (Feb 1, 2026)

Have a story we should cover? Contact us.