OpenAI cuts off Cursor after SpaceX acquisition, EU AI Act enforcement begins
OpenAI ends its model contract with Cursor after SpaceX buys the coding tool. Meanwhile, the EU AI Act started being enforced in August, and Google DeepMind's Co-Scientist can now run real lab equipment.
Every line links to a primary source in the full briefing.
01OpenAI ends its model contract with Cursor after SpaceX acquisition
OpenAI has announced it will wind down the contract under which it supplies models to Cursor, the popular AI coding assistant. The decision follows Cursor's acquisition by SpaceX. OpenAI has not provided a detailed public explanation beyond citing the acquisition as the reason.
If your team uses Cursor and relies on OpenAI models inside it, expect disruption: the tool may switch to different underlying models or degrade in capability, so it is worth testing alternatives now rather than being caught off guard.
Source: OpenAI
02The EU AI Act is now being enforced, and transparency rules are already live
From 2 August 2026, the European Commission's AI Office and national authorities began enforcing the AI Act. Alongside enforcement, new transparency requirements came into effect: certain AI systems must now inform users when they are interacting with AI rather than a human.
If you operate in Europe and deploy any AI-powered chat, voice assistant, or automated decision tool that faces customers, you are now legally required to disclose it is AI, and regulators are actively checking compliance.
Source: European Commission
03Google DeepMind's Co-Scientist moves from hypothesis writing to running actual experiments
Google DeepMind has expanded Co-Scientist, its Gemini-based multi-agent research system, well beyond generating scientific hypotheses. It now plans experiments, operates laboratory equipment, and produces scientific papers. Results have been experimentally validated across three disciplines, including materials synthesis and medical AI development.
This signals that AI agents are starting to close the loop in knowledge-intensive work, not just suggesting ideas but executing them, which will accelerate how quickly AI-driven competitors in research-heavy sectors can move.
Source: The Decoder
04Anthropic is wiring Claude agents directly into life sciences laboratories
Anthropic is pushing Claude beyond text tasks and into direct integration with life sciences R&D workflows. The goal is to have AI agents that can operate within laboratory environments and manage parts of the research process autonomously.
For businesses in pharma, biotech, or any research-heavy sector, this is an early signal of where autonomous AI agents are heading: not just assisting researchers but operating as participants in the workflow.
Source: R&D World
05US federal court rules Pentagon's supply-risk label on Anthropic is illegal
A US federal judge has ruled that the Department of Defense's designation of Anthropic as a supply-chain risk is unlawful. The ruling was reported by government technology outlets and appears to be a significant legal setback for the DOD's position.
The ruling limits the government's ability to restrict which AI vendors agencies can use, which could open doors for Anthropic to pursue more federal contracts and signals that courts are willing to push back on broad national-security framing of AI procurement decisions.
Source: MeriTalk
06Open-weight AI companies become the most sought-after acquisition targets in Silicon Valley
Capital is flowing heavily into AI companies that release their model weights publicly, making them the hottest acquisition targets in the current funding environment. The trend reflects large players wanting to own the open-weight ecosystem rather than just compete with it.
If you build on open-source or open-weight models, the companies behind those models may be absorbed by larger players, which could shift licensing terms, support, and the direction of future development.
Source: TechCrunch
Sources
6 links- 01Open-weight AI companies are the Valley’s hottest acquisition targetsTechCrunch
- 02Anthropic wants Claude to run life sciences R&D. Now it is wiring AI agents into the lab.R&D World
- 03Our decision on Cursor following its acquisition by SpaceXOpenAI
- 04Federal Judge Rules DOD Anthropic Supply Risk Designation Is IllegalMeriTalk
- 05Commission starts enforcing AI Act rules and new transparency requirements on 2 AugustEuropean Commission
- 06Google Deepmind's AI Co-Scientist now plans experiments, runs lab equipment, and writes scientific papersThe Decoder