Welcome to Issue 159 of The CTO Show Brief.
Infrastructure Is the New Software
The week ending February 22, 2026 confirmed what insiders have been pricing in for months: the locus of value creation has migrated from software abstraction to physical infrastructure. Sovereign nations are no longer passive capital allocators. They are direct owners of compute, energy, and aerospace assets. Meanwhile, the seat-based SaaS model took its most decisive blow yet, and the question of who governs AI in America escalated from policy debate into federal litigation.
Sovereign Wealth Is Buying the Stack, Not the Returns
WHAT CHANGED: MGX backed Anthropic's $30B Series G at a $380B valuation. Saudi Arabia's HUMAIN put $3B into xAI immediately before its SpaceX merger, converting AI equity into aerospace and satellite exposure. Qatar's QIA expanded its Fund of Funds to $3B, requiring VC firms to physically open offices in Doha.
WHY IT MATTERS: These are not financial bets. They are industrial policy executed through private markets. Gulf SWFs are acquiring stakes in the full AI stack: models, compute, satellite infrastructure, and venture pipelines. Capital is being used to manufacture geopolitical leverage in the intelligence economy.
WHO GAINS LEVERAGE: Founders building in regulated verticals or cross-border infrastructure who can credibly offer sovereign data assurance. The 'access to Gulf capital' story has shifted from relationships to strategic alignment on data sovereignty.
The Neo-Cloud Model Is Now Validated at Scale
WHAT CHANGED: Mumbai-based Neysa raised $1.2B (half equity, half debt via Blackstone) to deploy 20,000+ GPUs across India. The thesis is dedicated sovereign compute with faster rollout than hyperscalers, targeted at financial services and healthcare where data residency is non-negotiable.
WHY IT MATTERS: India currently has roughly 60,000 GPUs of capacity against a market Blackstone believes could reach 2 million units. Neysa is the template for the 'picks and shovels' phase in high-growth, data-sensitive markets. The neo-cloud model wins where compliance beats convenience.
WHO GAINS LEVERAGE: Infrastructure operators who can combine GPU access with regulatory compliance frameworks in markets like India, Saudi Arabia, and the UAE. Hyperscalers lose on speed and sovereignty. Startups filling that gap are now fundable at institutional scale.
SaaS Multiples Collapsed. The Cause Is Structural, Not Cyclical
WHAT CHANGED: Forward earnings multiples for enterprise software fell from 39x to 21x in weeks, driven by corporate procurement departments announcing plans to reduce software seats citing agentic AI efficiency gains. Anthropic's Claude Cowork, a desktop-native agent, was the proximate trigger.
WHY IT MATTERS: The seat-based subscription model assumed human growth as the revenue driver. That assumption is broken. Every agent that replaces a licensed seat removes a unit of software revenue permanently. This is not a correction. It is a structural re-rating of a business model.
WHO GAINS LEVERAGE: Vertical AI firms with deep workflow integration in regulated industries. Legora (legal AI) is reportedly raising at a $6B valuation after tripling from four months ago. Specialized, high-stakes, compliance-dense software retains pricing power. General-purpose horizontal SaaS does not.
AI Governance Moved from Policy to Federal Litigation
WHAT CHANGED: The DOJ's AI Litigation Task Force, established in January 2026, is expected to file its first lawsuits this month targeting Colorado's AI Act. The federal government is using executive authority to preempt state AI regulation before Congress acts.
WHY IT MATTERS: If the DOJ succeeds, the US gets a single national low-burden standard for AI compliance. If it fails, every AI lab is forced to maintain jurisdiction-specific model behaviors and safety protocols. The second scenario is a significant 'compliance tax' that advantages large incumbents and disadvantages every startup below Series B.
WHO GAINS LEVERAGE: Large AI labs, who can absorb compliance costs. Compliance and AI governance tooling vendors, who benefit either way. Smaller AI startups need to watch this closely. A fragmented regulatory outcome could be more disruptive than any model release.
Energy Is Now the Binding Constraint on AI Scale
WHAT CHANGED: Hyperscaler projected capex for 2026 hit $700B, up 60%. Amazon alone targets $200B in spend. Virginia data centers consume 26% of regional electricity. Ireland is approaching 32%. Multi-year backlogs for power infrastructure are forcing project delays and cancellations.
WHY IT MATTERS: The constraint on AI scale is no longer algorithmic or financial. It is physical. The companies that secure power purchase agreements, nuclear contracts, or on-site generation will build moats that cannot be replicated by writing a bigger check. Capital is now allocating toward energy, not just compute.
WHO GAINS LEVERAGE: Regions with surplus clean energy and stable grids. Infrastructure developers and energy technology companies, particularly those building SMRs or securing long-term PPAs. Founders in MEA benefit here: Gulf states with sovereign energy assets are well-positioned to offer compute with embedded power guarantees.
MEA Signal: Sovereign Capital Is Manufacturing Ecosystems, Not Just Funding Them
WHAT CHANGED: Qatar's QIA moved beyond passive LP investing, requiring international VC firms to establish physical Doha presence in exchange for Fund of Funds capital. Saudi Arabia's SVC launched an AI-powered market intelligence platform. Abu Dhabi's Origen secured $50M to deploy AI across government services and manufacturing.
WHY IT MATTERS: The Gulf has shifted strategy from deploying capital externally to importing institutional knowledge and manufacturing local ecosystems. This is the most significant long-term signal for MEA founders: the region is not just a source of LP capital anymore. It is building the infrastructure to keep value onshore.
WHO GAINS LEVERAGE: Regional founders who can anchor to government-aligned mandates. International firms with genuine MENA GTM capability. The 'parachute in for the check' approach to Gulf fundraising will become less viable as local ecosystem depth increases.
The open question for the next 12 months: As sovereign compute becomes a distinct asset class and software multiples re-price, where does application-layer talent migrate? The answer will define the next generation of high-value startups.
🎙️Episodes Recap:
AI is no longer just about models, prompts, or experimentation. It is becoming infrastructure. In this episode, I sit down with Shashank Tiwari , CEO and Founder of Uno.ai, to unpack one of the biggest shifts happening right now: AI is rapidly commoditizing, and the real value is moving up the stack. We explore how enterprises are moving from hype to real ROI, why AI agents introduce new risks, and how governance, control, and reliability are becoming critical in the age of autonomous systems. This conversation goes beyond the noise to focus on what actually matters for builders, operators, and investors.
As AI continues to automate workflows, decision-making, and even communication, one critical capability is becoming more valuable than ever: human connection. In this episode, Mehmet sits down with Joshua Bernstein , author of The Age of Synchrony, to explore the science behind trust, connection, and communication in an AI-driven world. They dive into the concept of synchrony, the neuroscience of human interaction, and why the ability to connect, build trust, and read people may become the ultimate competitive advantage for founders, leaders, and operators. From pitching investors to leading teams, this conversation explores how trust is built, why most decisions are emotional, and what happens to human purpose in a world where AI can do most of the work.
📖 From Nowhere to Next
Every week I share startup lessons and stories through The CTO Show Brief. But if you want to go deeper, my book From Nowhere to Next brings together the experiences and insights that shaped my own journey.
Thanks for reading — and for being part of this growing, global-minded network.
— Mehmet
