While much of the tech world was on holiday, dealmakers at Nvidia and SoftBank were busy making moves that would define the year in AI. In a quiet moment between Christmas and New Year’s, SoftBank announced it would acquire digital infrastructure investor DigitalBridge for approximately $4 billion—a strategic play to expand its footprint in data center and cloud infrastructure. Nvidia’s move, announced on Christmas Eve, was even more unusual. The company revealed a “non-exclusive licensing agreement” with Groq, the AI chip startup known for its custom hardware designed to accelerate AI inference—the real-time processing behind chatbot responses. Though not labeled an acquisition, the deal had all the hallmarks of one: Groq’s founder and CEO, Jonathan Ross, along with key engineering talent, will join Nvidia. Groq, which was valued at $6.9 billion just three months prior, will continue to operate independently. This pattern—acquiring talent and technology while leaving the startup intact—became a defining trend of 2025. It reflected a new era of dealmaking where the real prize wasn’t the company itself, but the people, IP, and strategic positioning it offered. The year’s most significant AI deals weren’t just about cash or ownership. They were complex, multifaceted arrangements that reshaped the industry’s power dynamics. Money was tied to access to compute. Licensing agreements came with talent transfers. Infrastructure deals locked governments and enterprises into long-term commitments. These weren’t traditional mergers—they were strategic power plays that redefined the landscape of Silicon Valley and beyond. Ranked by dollar value, these eight deals stood out as pivotal moments in the AI revolution: Stargate’s $500 billion infrastructure push The ambitious project to build a vast network of AI-powered data centers across the U.S. marked the largest infrastructure investment in tech history. Stargate’s Texas campus, designed to host thousands of AI servers, signaled a new era of compute scarcity and national strategic importance. Google’s $32 billion mega-buy Google’s acquisition of Wiz, a leading cloud security platform, was the most expensive AI-related purchase of the year. The deal aimed to secure Google Cloud’s position in the AI era by integrating advanced threat detection with its growing AI infrastructure. Meta’s $14 billion hire Meta’s investment in Scale AI—valued at $29 billion—was less an acquisition and more a strategic talent acquisition. CEO Alexandr Wang left Scale to join Meta’s AI leadership team, bringing with him deep expertise in data labeling and model training. The deal underscored Meta’s urgency to close the gap with OpenAI and Anthropic. Trump’s $8.9 billion Intel stake Former President Donald Trump’s undisclosed investment in Intel, facilitated through a private equity vehicle, gave him a major stake in the company. The move was widely seen as a political and technological power play, aiming to bolster U.S. semiconductor independence amid global AI competition. Google’s $2.4 billion cautionary tale Google’s purchase of Windsurf, a startup focused on AI-driven content moderation, ended in controversy after internal audits revealed ethical concerns. The acquisition highlighted the risks of rushing into AI deals without proper oversight. Disney’s $1 billion gamble Disney’s investment in OpenAI aimed to integrate generative AI into its content creation pipeline. The move was a bet on AI to transform storytelling, from scriptwriting to animation, and marked a major shift in how media companies are approaching AI. Nvidia’s Groq invasion The Christmas Eve deal with Groq, while not a full acquisition, was a masterstroke in talent and technology capture. By bringing in Ross and his team, Nvidia gained access to cutting-edge inference chip designs without the burden of integration. SoftBank’s $4 billion infrastructure play The acquisition of DigitalBridge was a calculated move to expand SoftBank’s global data center network. The deal gave the company a stronger foothold in the infrastructure layer that underpins AI scalability. These deals didn’t just move money—they moved power. In 2025, the future of AI wasn’t decided in labs alone, but in boardrooms, backrooms, and late-night negotiations.