← Back to newsletters

AI Daily Brief — Fri Apr 10

2026-04-10

By Vadym · Generated with AI, curated by me


Listen to this issue

AI governance met First Amendment law in a Colorado courthouse today. xAI filed suit to block state-level AI anti-discrimination rules while NVIDIA shipped physical AI reasoning models and robotics partners debuted new machines for National Robotics Week. Q1 2026 closed as the largest venture quarter in history — 80% of all global startup capital went to AI companies.


Headlines & News
Policy

xAI Sues Colorado to Block AI Anti-Discrimination Law — First Amendment Challenge That Could Cap All State AI Regulation

xAI filed suit in U.S. District Court on April 10 to block Colorado’s Senate Bill 24-205, which takes effect June 30. The law requires developers of “high-risk” AI systems to establish safeguards against discrimination in employment, housing, education, healthcare, and financial services, and mandates user disclosure when AI makes high-stakes decisions. xAI argues the law violates the First Amendment by compelling speech on contentious public issues and forcing Grok to reflect the state’s views on diversity rather than being objective. The company previously sued California over AI transparency laws and is seeking a court declaration that the Colorado law is unconstitutional and an injunction blocking enforcement before the June 30 effective date.

Robotics

NVIDIA Ships New Physical AI Models for National Robotics Week — Boston Dynamics, Caterpillar, LG Unveil Robots Built on Cosmos

NVIDIA released Cosmos Reason 2, a reasoning vision-language model that helps robots see, understand, and act in physical environments, alongside Cosmos Predict 2.5 and Cosmos Transfer 2.5 — open world models for large-scale synthetic data generation and simulation-based robot policy evaluation. NVIDIA also released Isaac Lab-Arena, a new benchmark for evaluating generalist robot policies in simulation before real-world deployment. Partners Boston Dynamics, Caterpillar, Franka Robots, LG Electronics, and NEURA Robotics debuted next-generation robots built on NVIDIA technologies, timed to National Robotics Week (April 6–12).

Industry

Perplexity Launches “Billion Dollar Build” — 8-Week Competition to Fund AI Startups Built on Its Computer Platform

Perplexity announced an 8-week competition where teams must build a company with a path to $1B using Perplexity Computer. Registration opens April 14 and requires an active Pro or Max subscription; entries are accepted until June 2. The top 10 finalists pitch live on June 9, with up to three winners announced June 10, each receiving up to $1M in seed funding from the Perplexity Fund and up to $1M in Perplexity Computer credits. Open to U.S. residents only, competing individually or in two-person teams.

Healthcare

Maine Sends AI Therapy Chatbot Ban to Governor — 98 AI Bills Now Moving Across 34 States

Maine lawmakers approved LD 2082, which bans AI from clinical roles in mental health therapy while permitting administrative use. The bill now goes to the governor for signature. Missouri’s HB 2372 — an omnibus health care bill including a therapy chatbot ban with $10,000 first-violation penalties — cleared the full House and now sits with the Senate Committee on Families, Seniors and Health. Nationally, the Future of Privacy Forum is tracking 98 chatbot-specific bills across 34 states and three federal proposals.

Funding

Q1 2026 Shatters VC Records — $300B Raised Globally, 80% to AI; Four Companies Captured 65% of All Startup Capital

Q1 2026 set an all-time record for global venture capital: $300 billion invested across 6,000 startups, up 150%+ year-over-year. AI companies attracted $242 billion — 80% of all global VC. Four mega-rounds dominated: OpenAI ($122B), Anthropic ($30B), xAI ($20B), and Waymo ($16B), totaling $188 billion, or 65% of all global startup investment for the quarter. Late-stage funding reached $246.6 billion, up 205% year-over-year across 584 deals. U.S. companies captured 83% ($250B) of total global VC; China was second with $16.1B, the U.K. third with $7.4B.

OpenSource

Alibaba Ships Qwen 3.6-Plus — 1M-Token Context, Always-On Reasoning, Targets Western Enterprise Codebases

Alibaba released Qwen 3.6-Plus, the latest in its flagship model series, removing the thinking/non-thinking toggle from the 3.5 generation — reasoning is now always active on every prompt. The model features a 1-million-token context window by default, a hybrid linear attention plus sparse mixture-of-experts architecture, and strong multimodal reasoning including high-density document parsing, physical-world visual analysis, and long-form video reasoning. Available via Alibaba Cloud’s Model Studio and Qwen Chat; compatible with major third-party coding assistants.


Analysis

Takeaway

Today’s stories land on two fault lines that will define the next few years: who governs AI, and who can build with it. xAI’s Colorado lawsuit signals that state-level AI regulation will face First Amendment challenges before it can fully take hold — while Maine’s therapy chatbot ban shows the opposite pressure: narrow, emotionally clear restrictions moving fast across 34 states. Meanwhile, NVIDIA is building the simulation infrastructure for a world full of physical robots, Alibaba shipped a frontier model with always-on reasoning targeting enterprise codebases, Q1 2026 closed as the largest venture quarter in history with 80% of capital going to AI, and Perplexity launched a competition to prove its computer platform can build billion-dollar companies in 8 weeks. The capital and the tools are concentrated. The legal framework for what those tools can do is a battlefield.

— Boba


Curated by Vadym