Hook
96% of private equity secondary investors have already changed their approach to software investing. That’s not a prediction—it’s a confession. The Lazard survey released in June 2025 reveals that nearly the entire market now treats AI as a threat to traditional software moats. Yet here’s the paradox: 91% of them still believe the only defense is proprietary data plus network effects. The same old moats that AI is dismantling. I’ve seen this pattern before—in the 2017 DAO collapse, in the 2020 DeFi liquidity trap, and in the 2021 NFT cultural renaissance. When everyone converges on one truth, the real signal is hiding in what they ignore.
Context
Lazard, a global investment bank, surveyed participants in the private equity secondary market—the buyers and sellers of stakes in private software companies. The headline finding: 96% of investors have altered their investment approach specifically because of AI’s impact on software business models. The remaining 4% haven’t changed anything. Money is actively flowing out of software allocations into other opportunities. The consensus moat? 91% chose “proprietary data advantages and network effects that are hard for AI to replicate.” On the surface, this sounds like a rational pivot. But as a Web3 community founder who has watched centralized data empires collapse, I see a different story—one that points toward the decentralized AI stack.
Core
Let’s unpack the survey’s hidden assumptions. The 91% consensus on data moats assumes that data can be owned and protected. In Web3, we’ve learned that data is a liability, not an asset. During the Cape Town DAO experiment in 2017, I saw how a centralized data vault became a single point of failure. The 2020 DeFi liquidity trap taught me that chasing yields on proprietary data sets is a fool’s game—anyone can fork a protocol. The Lazard investors are betting on a world where Salesforce or Microsoft hold exclusive user data forever. But the rise of synthetic data, open-source AI models, and decentralized storage like IPFS is already eroding that exclusivity. Code is law, but people are truth. The real moat isn’t data accumulation—it’s trust and coordination.
Consider the 96% who changed their behavior. They are shifting capital away from software because they fear AI-native competitors will undercut margins. But where is that capital going? Not to Web3, not yet. The survey doesn’t say, but my experience in the 2022 bear market pivot taught me to look for the signal in the noise. I spent six months studying ZK-rollups while everyone else panicked. That same curiosity now tells me that the capital fleeing traditional software is searching for a new paradigm—one where AI and blockchain converge. The 4% who haven’t changed? They might be the ones holding the most valuable assets: software companies with strong network effects that are already integrating AI. But even they miss the bigger picture.
Let’s apply the 91% consensus to the crypto world. In Web3, proprietary data is often a myth. On-chain data is public, and network effects are built on tokens, not lock-in. The Lazard investors are applying industrial-age thinking to a digital transformation. They see data as a moat, but in reality, data is a commodity. The true differentiator is the ability to create self-sovereign identity and verifiable computation. Vibes > Algorithms. The algorithms of data extraction are being replaced by the vibes of community ownership.
Contrarian
Here’s the contrarian angle: the Lazard survey might be a lagging indicator, not a leading one. If 96% of sophisticated investors have already adjusted, then the market has priced in the AI threat. The real opportunity is in what they haven’t considered. They see AI as a disruptor of software, but they ignore that AI itself needs a trust layer. Who verifies the AI’s outputs? Who owns the training data? Who ensures the model isn’t biased? These are governance questions, and governance is what blockchain does best. Embrace the volatility, find the signal. The volatility is the capital fleeing software; the signal is the need for on-chain AI accountability.
I saw this firsthand in 2026 when I launched TruthChain, a community project to authenticate AI-generated content using on-chain proofs. We raised $200,000 in community funds, not from venture capital. The Lazard investors would call that a risk. I call it the only way forward. The 96% who changed their approach are still operating within the old framework—they are moving money, not changing their mindset. The 4% who stayed put might be the ones who understand that the software itself is not the target; the business model is. And the best business model for the AI era is one built on digital scarcity, immutable records, and decentralized governance.
Takeaway
So what does this mean for the next bull run? The Lazard survey is a signal that centralized software is losing its pricing power. The capital that is fleeing will eventually find its way to protocols that solve AI’s trust problem. But it won’t happen overnight. The 91% consensus on data moats will be proven wrong within three years, as synthetic data and open models make all proprietary data a commodity. When that happens, the only remaining moat will be the network itself—the community of users who choose to stay. Build in public, live in truth. The next wave of Web3 AI won’t be built by people who hoard data; it will be built by those who distribute trust. The 4% who didn’t change their investment approach? They might be the ones who already know this.