AI dealmaking surges, from roll-ups to Anthropic's compute push — 5/7/2026

Key Concepts

• AI Rollup: A strategy where venture-backed firms acquire established, non-tech companies (e.g., in services, healthcare, construction) to modernize them by deploying proprietary AI platforms and engineering teams.
• Agentic Engineering: The evolution of software development where engineers no longer write code manually but instead manage AI agents that write, test, and deploy code.
• Compute Cycle: The massive capital expenditure (capex) cycle driven by the need for Nvidia GPUs and data center capacity to power AI models.
• Service-as-Software: The transition of traditional service-based businesses into high-margin, AI-driven entities that scale without linear headcount growth.
• Nexus Platform: A proprietary AI platform used by firms like Long Lake to provide specialized workflows for niche industries.
• Loops and Routines: Advanced AI features that allow agents to run tasks on schedules or in response to events, moving beyond simple chat-based interactions.

---

1. The AI Deal Landscape

The AI trade remains robust, characterized by a shift from simple software adoption to deep structural integration.

• Private Equity & AI Labs: Major AI labs are securing massive partnerships with private equity. OpenAI launched a $10 billion deployment vehicle with TPG, Brookfield, Advent, and Bain. Anthropic secured a $1.5 billion joint venture with Blackstone, Hellman & Friedman, and Goldman Sachs.
• Compute Infrastructure: Anthropic announced a deal with SpaceX to utilize the "Colossus 1" data center in Memphis, adding 220,000 Nvidia GPUs and 300 megawatts of power. This complements existing massive compute partnerships with Amazon, Google, and Microsoft.

2. The "AI Rollup" Strategy

General Catalyst and its portfolio company, Long Lake, are pioneering a new M&A theme: taking public companies private to perform "AI surgery."

• The Playbook: Instead of selling AI tools to companies, they acquire the companies outright, embed applied AI engineers, and use their proprietary Nexus platform to optimize workflows.
• Case Study: Long Lake agreed to take American Express Global Business Travel private for $6.3 billion (a 60% premium).
• Economic Impact: The goal is to transform service-based businesses (which historically struggle to scale without adding headcount) into high-growth, high-margin entities. Long Lake claims to have increased revenue growth from 0–5% to over 20% in their acquired companies.

3. Morgan Stanley Founder Summit Insights

Mandel Crawley, Chief Client Officer at Morgan Stanley, shared findings from their inaugural Founder Summit research:

• The AI Gap: While 95% of founders believe AI is critical, only 23% feel supported in implementing it. This "teething process" involves the difficulty of operationalizing AI and managing the human change-management aspect.
• Liquidity Trends: Despite the abundance of private capital, two-thirds of founders still plan to IPO. However, the path to liquidity has diversified, with $20 billion in tender transactions facilitated by Morgan Stanley last year alone.
• Founder Sentiment: Founders view personal wealth and business success as a single, parallel path. The current environment is viewed as "constructive," with founders focusing on growth despite macroeconomic instability.

4. Anthropic and the Future of Software

Boris Churnney and Cat Woo (Cloud Code leads) discussed the shift toward Agentic Engineering.

• 100% AI-Generated Code: Anthropic reports that 100% of their internal "Cloud Code" is now written by Claude. Engineers have transitioned from writing code to managing agents that write code.
• Printing Press Moment: The team argues that AI is a "printing press moment" for software, democratizing the ability to build and innovate.
• Judgment and Taste: Contrary to the belief that AI lacks "taste," the team notes that Claude’s ability to push back on bad ideas and produce high-quality, production-ready code is improving rapidly.
• Safety vs. Speed: Anthropic emphasizes that safety is the primary constraint. They balance speed by requiring human confirmation for agentic actions, allowing users to build trust in the model's capabilities over time.

5. Synthesis and Conclusion

The current AI landscape is defined by a transition from "AI as a chatbot" to "AI as an operational engine."

• For Enterprises: The key takeaway is that AI cannot remain on the periphery; it must be at the center of business processes, similar to how computers were integrated in the 1990s.
• For Investors: The "AI Rollup" represents a new, potentially massive asset class that combines the stability of legacy businesses with the growth profile of tech companies.
• For Developers: The role of the software engineer is evolving into an "agent manager," where the primary skill is no longer syntax, but the judgment of what to build and how to orchestrate AI agents to execute that vision.

Notable Quote: "Software as a Service (SaaS) was the definition of the last era. I believe it’s going to be Service as Software." — Medu Nimbur, General Catalyst.