AI and the question of defensibility

Everyone talks about post-acquisition AI strategy. Fewer talk about AI defensibility. Could AI wipe out the competitive advantage you have just paid a premium for in six months?

Many tech-enabled businesses have proprietary systems that, done well, create real advantage and justify value. The optimised CRM inside a professional services firm. The smarter TMS in a logistics company. The niche booking engine refined over years.

In an AI era, how deep is that moat really? We are not talking about someone vibe-coding a competitor in a weekend. We are talking about a well-funded competitor leveraging AI-enhanced development tools to build something equivalent, perhaps better, in three months. The cost of reproducing software has fallen dramatically. The cost of reproducing customers, data and trust has not.

If that is possible, is the business still worth the same premium? Or has part of the advantage already been neutralised?

So where does defensibility actually live? Proprietary data. Switching costs. Embedded processes. Network effects. Ecosystem integration. Regulatory complexity. Distribution. In an AI era, value has to sit somewhere harder than code.

For acquirers, that is an extra lens for technology due diligence. The right question is no longer just "is the technology good?" It is "if a capable competitor rebuilt this with modern tooling, what would still be hard to copy?"