Crypto Briefing dropped a headline that screamed 'builds investment empire,' but the body was a desert. One fact survived: Wiz CEO Assaf Rappaport, fresh off the Google acquisition deal, is writing checks to AI cybersecurity startups. The article itself is thin—borderline clickbait, source mismatch, zero analysis. As a macro watcher who manages a digital asset fund, I've learned to ignore the packaging and extract the signal. The signal here is not the story Crypto Briefing tried to sell, but the capital flow behind it.
Let's rewind. Rappaport sold Wiz to Google for a reported $23 billion in 2026—one of the largest cyber acquisitions ever. Now he's deploying a portion of that liquidity into AI-native security companies. This is not a random act of charity. It's a structural bet on where the next billion-dollar attack surface lies. And that attack surface is directly relevant to blockchain infrastructure, even if the crypto press can't articulate why.
Context: The liquidity map has shifted. Post-ETF approval, Bitcoin became Wall Street's toy—Satoshi's vision of peer-to-peer cash is dead, replaced by a macro asset. Meanwhile, real innovation capital is flowing into AI and cloud security. The intersection of AI agents and decentralized verification is where I've been positioning my fund since 2025. Rappaport's move validates that thesis, but only if you look beyond the surface.
The core insight is simple: AI security needs trustless verification. Today, most AI inference runs on centralized cloud providers like AWS or Google Cloud. The audit logs, model integrity checks, and threat detection all depend on a single party's honesty. That's a single point of failure. As AI agents become autonomous and execute financial transactions on-chain, the demand for verifiable compute will explode. This is where blockchain—specifically zero-knowledge proofs and decentralized oracle networks—becomes essential.
Rappaport's investment targets are unconfirmed—Crypto Briefing failed to name them—but the pattern is clear. He's betting on companies that use AI to automate security operations (SOC automation, adversarial ML detection) and possibly those that integrate with Web3 identity or access control. The macro trend: as more enterprises move to multi-cloud and edge computing, the perimeter dissolves. AI-driven security becomes the only scalable defense. Blockchain's role is to provide a cryptographic root of trust for the AI's decisions.
Contrarian angle: most crypto natives will ignore this. They're still chasing memecoins or arguing about L2 data availability (which is overhyped—99% of rollups don't generate enough data to need dedicated DA). Meanwhile, institutional capital is flowing into AI security startups at a record pace. PitchBook data shows Q1 2027 AI security funding hit $4.2 billion, up 60% year-over-year. The disconnect between on-chain speculation and real-world infrastructure building grows wider. When the next bull cycle arrives, it won't be fueled by another DeFi summer—it will be driven by AI agents that need trustless payment rails and verifiable compute. Rappaport sees that. You should too.
Let's break down the mechanics. AI agents require micropayment channels to pay for API calls, storage, and compute. Current L2 solutions can handle a few hundred transactions per second, but agent-to-agent trading could require millions. This demands a new class of execution environments optimized for machine-to-machine value transfer. I've been tracking projects like Akash, Render, and a handful of private ZK-rollup teams building this infrastructure. Rappaport's portfolio likely overlaps.
Follow the gas, not the hype. The Gas here is not Ethereum gas, but the energy invested in solving real problems. Crypto Briefing's article is hype—low information density, designed to capture clicks. The real gas is the engineering talent moving from cloud security into decentralized verification. I've audited whitepapers since 2017; I can tell you when a project is built on marketing versus math. This trend is math-based. The cryptographic primitives—ZK-SNARKs, threshold signatures, secure enclaves—are ready for prime time. What's missing is the killer app. AI security might be it.
Bets are cheap; exits are expensive. Rappaport can place small bets on a dozen startups; that's cheap capital for him. The expensive move is exiting at the right time. For crypto investors, the same principle applies: buying AI-related tokens now may seem cheap, but many will zero out. The winning plays will be infrastructure layers that enable verifiable AI, not the AI apps themselves. I've structured my fund to hold long positions in decentralized compute networks and ZK-focused L2s, while avoiding the hype-driven AI agent token craze.
Here's a concrete example from my portfolio. In Q4 2026, I allocated $2 million to a startup building a decentralized inference verifier using recursive ZK proofs. They were incubated by a major cloud provider (non-disclosure prevents naming). The team came from Palo Alto Networks and had audited Wiz's security stack. Their pitch: any AI output can be cryptographically signed and verified by a smart contract, enabling autonomous DeFi loans based on AI risk assessments. That's the kind of infrastructure Rappaport probably backs.
Takeaway: Position for the convergence, but with a macro lens. The liquidity cycle is still bearish—central banks are tightening, and crypto markets are bleeding. Survival matters more than gains. Use the quiet period to study which protocols have real users paying fees. Check gas consumption on rollups; ignore TVL games. If a project claims to be the 'AI layer for crypto' but can't show weekly active agents, it's vaporware. Rappaport's investment is a macro signal, not a trade signal. It tells us where smart capital is looking. Our job is to triangulate the exact coordinates.
I'll leave you with a final thought: The AI-crypto convergence is not a narrative—it's an engineering necessity. When autonomous agents start negotiating with each other for compute resources, they need a settlement layer that is permissionless, borderless, and mathematically guaranteed. That layer will be blockchain. But the route will be ugly—false starts, scams, and overhyped VCs. Rappaport can afford to bet on 20 startups and win on two. For a fund manager, we need to bet on the protocol that becomes the TCP/IP for machine-to-machine value flows. That's the real empire building.
Follow the gas, not the hype. Bets are cheap; exits are expensive. Crypto markets are finally growing up, and the toys are being put away.
