动察 Beating AI News Flash: Cybersecurity giant Palo Alto Networks has announced plans to acquire AI startup Applied Compute for $4.5 billion in an all-stock deal. The company had just completed a $350 million funding round on November 2, reaching a valuation of $3.25 billion.
Applied Compute was founded by 3 former OpenAI researchers about 18 months ago and currently has approximately 70 employees. The company says its annualized revenue run rate has reached approximately $100 million.
Applied Compute specializes in helping enterprises train their own AI models. It builds on open-weight models, further trains them with enterprises' business data, and then handles deployment and continuous improvement. For example, its custom model for legal AI company Harvey improved answer and citation quality while cutting inference costs to less than half of a certain frontier model. It also trained a model routing system for Clay, reducing inference costs by up to 47%.
Palo Alto Networks originally focused on cybersecurity products and now plans to expand further into enterprise AI model training and operation, enabling customers to both use their own proprietary models and obtain accompanying security protection. Applied Compute founder Yash Patil will serve as co-chief AI officer after the deal closes.
According to the acquirer's disclosure, the $4.5 billion consideration includes approximately $500 million in cash and prepaid compute commitments. The transaction still requires regulatory approval and is expected to close between February and April 2027.
