Abu Dhabi-headquartered Shorooq has participated in the $1.03 billion seed round of AMI Labs, the artificial intelligence company founded by Turing Award winner and former Meta Chief AI Scientist Yann LeCun, in what ranks as one of the largest seed financings an AI startup has ever raised.
The round, which values AMI Labs at approximately $3.5 billion pre-money, was co-led by Cathay Innovation, Greycroft, Hiro Capital, HV Capital, and Bezos Expeditions, with participation from Temasek, Nvidia, SBVA, Mark Cuban, Eric Schmidt, and Xavier Niel.
AMI Labs is building what LeCun calls "world models," AI architectures designed to learn from spatial and real-world data rather than purely predicting text or images, with the goal of enabling systems that can reason about cause and effect and plan actions in physical environments.
The company is headquartered in Paris with offices in New York, Montreal, and Singapore, and will initially target enterprise applications across manufacturing, aerospace, robotics, and biomedical industries.
It’s led by CEO Alexandre LeBrun, formerly co-founder and CEO of healthcare AI company Nabla, with LeCun serving as Executive Chair.
The investment extends Shorooq's positioning as a conduit between global frontier AI and Gulf capital.
The firm's $100 million Presight-Shorooq Fund I, a joint venture with G42's ADX-listed AI arm, disclosed its first five investments in February of this year after screening more than 1,000 companies in 120 days, with co-investors across that portfolio including Andreessen Horowitz, Index Ventures, Google Ventures, and Y Combinator, and LeCun himself appearing as an angel investor.
Shorooq also launched a separate $200 million late-stage growth fund at Web Summit Qatar backed by QIA, taking the firm past $1 billion in assets under management across venture, credit, private equity, and real assets.
"AMI Labs represents one of the most ambitious efforts to redefine the foundations of artificial intelligence," said Dr. Bilal Baloch, Partner at Shorooq. "Yann LeCun's vision for world-model architectures could fundamentally expand what machines are capable of understanding and executing in the physical world."
