PE firms reassure investors as SaaS sell-off continues — but not all software moats are equal
Feb 11, 2026
Key Points
- PE firms including Thoma Bravo, Vista Equity, KKR, and Ares are reassuring investors that AI disruption won't harm their software portfolios, but their arguments contradict each other depending on what they own.
- Software defensibility increasingly depends on structural moats like network effects and supply networks rather than engineering cost, as AI tools lower the barrier to building feature-equivalent code.
- PE shops are avoiding the harder task of separating holdings with real competitive advantages from those that relied only on engineering scarcity, making valuations difficult to reset.
Summary
Private equity firms backing software portfolios are moving to reassure investors as SaaS valuations face pressure from AI disruption fears. Thoma Bravo held investor calls, Vista Equity's CEO sent a client email and scheduled a webinar, and leaders at KKR, Blackstone, and Ares told investors their exposure to software commoditization is limited and the sell-off is overblown.
The firms are making contradictory claims. Software-only PE shops say AI won't disrupt their assets. Diversified firms say their software exposure is small enough to ignore. Neither stance acknowledges what actually matters: not all software moats are equal.
The real divide isn't between software that will be disrupted and software that won't. It runs between software where defensibility comes from something concrete—network effects, liquidity, regulatory barriers, brand—and software where defensibility comes from engineering cost alone.
Airbnb could theoretically be cloned with AI tooling for a fraction of historical build cost. You would still have no houses, no users, no liquidity. The moat is the supply network, not the code. Duolingo has proven resilient partly because people love the product and brand, not just because the software is hard to replicate.
Chegg is the counterpoint. The core asset was a database of answers that could be scraped and fed into ChatGPT. Stock down 90 percent. The code was never the moat.
As AI lowers the cost to build equivalent software from $1 billion to $100 million to $10 million, PE firms will need to separate companies with real structural defensibility from those that only had engineering scarcity protecting them. The messaging from the big PE shops right now skips that distinction. It's easier to tell investors AI won't hurt the portfolio than to revalue every holding and admit some of them look like Chegg, not Airbnb.