AI News Roundup: Google I/O 2026 Keynote, Anthropic Acquires Stainless, EU Streamlines AI Act

Google I/O 2026 Keynote Arrives with Gemini Upgrades and Android XR Glasses

Google I/O 2026 opens today at the Shoreline Amphitheatre in Mountain View, with CEO Sundar Pichai delivering a keynote at 1 p.m. ET. The centerpiece is expected to be a major Gemini model upgrade — whether it lands as Gemini 4.0 or an advanced 3.x iteration, it will position Google to compete directly with OpenAI’s GPT-5.5. Pre-event leaks revealed seven hidden Gemini Live voice models under internal testing, plus references to Gemini Omni, a unified pipeline for generating text, images, and video.

Hardware steals the show too. Google is previewing two flavors of Android XR glasses built in partnership with Samsung, Gentle Monster, and Warby Parker — one display-free pair focused on hands-free Gemini interaction, and a more ambitious model with an in-lens display for navigation and translation overlays. Also expected: Android 17 with deep “Gemini Intelligence” integration, the new Googlebook premium laptop category, and expanded agentic AI capabilities via Google’s autonomous assistant Remy.

Anthropic Acquires Stainless for $300M+, Cutting Off Rivals’ SDK Pipeline

Anthropic announced yesterday it has acquired Stainless, the developer tools startup founded by former Stripe engineer Alex Rattray, for more than $300 million. Stainless generates SDKs, CLIs, and MCP servers used by hundreds of companies — including Anthropic’s own rivals OpenAI, Google, and Cloudflare. The deal gives Anthropic control of a critical piece of developer infrastructure while taking it away from competitors.

Anthropic said it will wind down all hosted Stainless products, though existing customers retain full ownership of SDKs they’ve already generated. This marks Anthropic’s fourth acquisition in six months, following Bun (December 2025), Vercept (February 2026), and Coefficient Bio (April 2026). The acquisition aligns with Anthropic’s push to own more of the AI development stack as it builds out its agentic ecosystem.

Anthropic Eyes $950B Valuation in Blockbuster Fundraise

In parallel with its acquisition spree, Anthropic is negotiating a $30–50 billion raise at a valuation of up to $950 billion, according to Bloomberg and the New York Times. If completed, the round would make Anthropic the world’s most valuable startup, surpassing OpenAI’s $852 billion valuation from March. The round is expected to close by month’s end.

The valuation surge is backed by explosive growth: Anthropic’s annualized revenue has climbed past $30 billion, up from roughly $9 billion at the end of 2025. For the first time, more American businesses are paying for Claude than for ChatGPT — 34.4% adoption versus 32.3% for OpenAI in April, according to VentureBeat. The company has also held preliminary IPO discussions, with a potential listing eyed for late 2026.

EU Council and Parliament Agree to Streamline the AI Act

The European Council and Parliament reached a provisional agreement on May 7 to simplify key provisions of the EU AI Act, marking the first set of amendments since its adoption in June 2024. High-risk AI system obligations have been pushed back 16 months — from August 2026 to December 2027 — giving businesses significantly more runway to comply.

The amendments also introduce two new prohibited practices: using AI to generate non-consensual intimate imagery and child sexual abuse material. On the business-friendly side, regulatory exemptions previously reserved for SMEs now extend to small mid-cap companies, and the AI Office’s enforcement powers have been reinforced. The text is expected to receive formal adoption by July 2026.

Colorado and Connecticut Reshape U.S. State AI Law

Colorado Governor Jared Polis signed SB 189 on May 14, substantially overhauling the state’s landmark AI Act. The revised law pivots away from its previous prescriptive, risk-based framework toward a more limited, disclosure-focused model — a pragmatic acknowledgement that enforcement of risk classifications proved impractical. Meanwhile, Connecticut’s bipartisan SB 5 passed the state legislature on May 1 and is expected to be signed by Governor Lamont, adding another state to the growing patchwork of AI regulations.

These state-level moves come as the Trump Administration, once staunchly opposed to AI regulation, is now considering oversight for advanced models after Anthropic’s “Mythos” system demonstrated the ability to identify critical infrastructure vulnerabilities at scale. All five major frontier AI labs now participate in evaluations through the federal CAISI framework.

SpaceX Secures $60B Option to Acquire Cursor

SpaceX announced a deal structure giving it the right to acquire AI coding startup Cursor for $60 billion later this year, or alternatively pay $10 billion for a partnership arrangement. The deal follows Elon Musk’s February merger of SpaceX and xAI, valued at $1.25 trillion, as part of his plan to transform the rocket company into an AI powerhouse ahead of its upcoming IPO.

Cursor’s valuation trajectory has been extraordinary: $2.5 billion in January 2025, $9 billion by May 2025, $29.3 billion after its Series D in November, and now a potential $60 billion exit. The startup is simultaneously in talks to raise another $2 billion at a $50 billion+ valuation, keeping its options open.