Advanced Machine Intelligence (AMI), a new Paris-based startup cofounded by Yann LeCun, has secured massive funding to build AI systems that understand the physical world. According to multiple reports, AMI raised €890 million in funding from investors including Nvidia, Samsung, Toyota, Eric Schmidt, and Jeff Bezos. However, some sources report the amount as more than $1 billion, a discrepancy that may stem from currency conversion—€890 million is approximately $970 million. The company aims to develop 'world models' based on the JEPA project, which LeCun has described as necessary for human-level intelligence. According to WIRED, LeCun described that most human reasoning is grounded in the physical world, not language, and that extending large language models to human-level intelligence is 'complete nonsense'. AMI will have offices in Paris, Montreal, Singapore, and New York, and LeCun will continue working as a New York University professor. The startup is LeCun's first commercial endeavor since his departure from Meta in November 2025. AMI was cofounded by LeCun, Michael Rabbat, Laurent Solly, Pascale Fung, Alexandre LeBrun, and Saining Xie. The company says in a press release that it aims to build AI systems that understand the world, have persistent memory, can reason and plan, and are controllable and safe. LeCun has stated that AMI aims to work with companies in manufacturing, biomedical, robotics, and other industries, and could build a realistic world model of an aircraft engine to help optimize efficiency, minimize emissions, or ensure reliability.
The exact funding amount and investor list for AMI contain contradictions. While major media outlets report that AMI secured €890 million from corporate investors including Nvidia, Samsung, Toyota, Eric Schmidt, and Jeff Bezos, according to WIRED, the round was co-led by venture firms Cathay Innovation, Greycroft, Hiro Capital, HV Capital, and Bezos Expeditions, with additional backers including Mark Cuban, Eric Schmidt, and Xavier Niel. The discrepancy may reflect different stages or tranches of the round, but the exact total in a single currency and the complete investor list remain unclear.
AI agents generate software so fast that it's nearly impossible for humans to review it all.
In other European AI funding news, Mistral has raised billions for AI infrastructure. According to multiple reports, Mistral raised $830 million (€750 million) to build a data centre near Paris running on over 13,000 NVIDIA chips with 44 MW compute power. The company plans to build 200 MW of compute capacity across Europe by next year. Additionally, Mistral announced a $1.4 billion (€1.2 billion) investment in Sweden for AI infrastructure, signaling a major push to expand European AI capabilities.
Anysphere, the company behind the AI coding tool Cursor, is in talks to raise about $2 billion at a valuation over $50 billion, according to CNBC and Bloomberg News citing sources. Forbes reports that Anysphere's co-founder Arvid Lunnemark from Malmö owns about 4.5% of the company, worth around 12 billion SEK before the new round. The massive valuation reflects the growing demand for AI-powered developer tools.
Most human reasoning is grounded in the physical world, not language, and that AI world models are necessary for human-level intelligence.
Cloudsmith, a software supply chain platform, announced a $72 million Series C financing led by TCV with participation from Insight Partners and other existing investors. The company's CEO, Glenn Weinstein, said in a press conference that AI agents generate software so fast that it's nearly impossible for humans to review it all, highlighting the need for automated security and management solutions.
Synera, an engineering automation platform, announced a $40 million (€35 million) Series B round led by Revaia, with participation from Capgemini through ISAI Cap Venture and existing investors. Synera's platform integrates with more than 80 engineering tools and enterprise systems, enabling companies to automate complex engineering workflows.
Extending LLMs to human-level intelligence is 'complete nonsense'.
Waiv, a medical imaging AI company, announced a $33 million financing round co-led by OTB Ventures and Alpha Intelligence Capital, with participation from others. Waiv's products include RlapsRisk BC, MSIntuit Suite, and BRCAura, which use AI to analyze medical images for early detection and diagnosis.
Halter, a New Zealand-based agritech company, is in talks to raise fresh funding led by Founders Fund, with the round heavily oversubscribed. Halter's system uses solar-powered collars that provide real-time data on location and health indicators, and emit sound and vibration signals to guide livestock. The company's technology aims to improve farm efficiency and animal welfare.
Alva Labs, a Swedish AI recruitment platform, is launching an AI-driven recruitment system and raising €20 million through a convertible loan. Founded by three EQT veterans, Alva Labs has revenues of nearly 90 million SEK. The new system aims to automate candidate screening and assessment using AI.
A London-based business raised £2 million to develop an AI-powered coffee vending machine that tailors drinks based on customer preferences. According to City AM, a columnist expressed skepticism about the need for AI in coffee machines and noted that AI is being added to products just for the sake of it.
According to Gartner's 2025 CIO survey, 86% of manufacturing respondents will increase investment in generative AI in 2026, with 86% expected to deploy GenAI by 2026 and 97% by 2028. However, a 2024 Gartner survey found that manufacturing CIOs report that only an average of 41% of AI and GenAI prototypes reach production, indicating a significant gap between experimentation and deployment.