Alongside data, proprietary knowledge is emerging as a major source of defensibility.
Many organisations are sitting on valuable knowledge assets hidden across policies, documents, service interactions, internal playbooks, support histories and expert workflows. The opportunity is not simply to store that knowledge, but to structure it and apply it in useful contexts.
That is where AI can create outsized value.
Businesses that can turn fragmented knowledge into embedded capability — inside products, workflows and decision-making — are doing more than adding AI features. They are building differentiated intellectual property.
For investors, this matters. A business with real proprietary context is in a very different position from one simply layering a model on top of publicly available information.
The biggest opportunity is not blanket automation
One of the more useful ways to think about AI is that it is replacing tasks, not replacing entire jobs. That distinction matters.
Most roles consist of a mix of repetitive work, analytical work, judgment, communication and accountability. AI is already proving useful in reducing lower-value, time-consuming work. It is much less effective at replacing the context, responsibility and judgment that sit around it.
This is why the best AI adoption stories are often more grounded than the headlines suggest.
The real gains are showing up in:
- Faster research and analysis
- Shorter turnaround times
- Better workflow automation
- Improved decision support
- More responsive customer experiences
- More productive engineering teams
That is valuable because it removes toil and frees people to focus on higher-order work. But it is not the same as replacing the human layer altogether.
For private equity-backed businesses, the implication is clear: the goal should be to redesign work around leverage, not chase simplistic narratives about headcount elimination.
AI works best when embedded in real workflows
The most effective uses of AI are rarely the most theatrical.
In many businesses, the strongest results are not coming from broad attempts to automate the entire organisation. They are coming from AI embedded inside vertical software, existing workflows and domain-specific tasks.
That might mean:
- Enriching CRM data through call intelligence
- Surfacing insights inside contract or compliance workflows
- Improving service operations with contextual support intelligence
- Accelerating delivery within software engineering pipelines
- Using domain-specific copilots grounded in internal systems and knowledge
This is an important point for investors and operators alike. The market often talks about AI as if value is created through general-purpose capability alone. In reality, much of the most practical value is being created through focused 'AI injections' into tools and processes businesses already rely on.
That is often where competitive advantage becomes tangible.
Why customer intimacy matters more, not less
As AI lowers the cost of building features, understanding customers becomes even more strategic.
Businesses that know their users deeply, fit solutions into actual workflows, and solve real pain points are better positioned to defend value than those relying on generic capability. Large platforms may have scale and broad tooling, but they do not automatically have the same proximity to the customer problem.
That is why old-school business fundamentals still matter.
Customer intimacy, trusted relationships, responsiveness and service quality are not outdated in the AI era. In many markets, they are exactly what prevent commoditisation.
For portfolio companies, this should be reassuring. AI may change how products are delivered and how teams operate, but it does not eliminate the advantage of being closer to the customer than larger, less specialised competitors.
Guardrails are now a value protection issue
If AI creates new opportunities, it also creates new forms of risk.
For many businesses, especially those in regulated or high-stakes environments, the biggest challenge is not whether AI can automate something. It is whether that automation can be trusted, understood and governed.
This is particularly relevant when businesses start using agentic models or more autonomous workflows. Autonomy sounds attractive, but black-box behaviour introduces real risk. If a company cannot explain how an AI system reached an answer, what checks were performed, or where accountability sits, that becomes a governance problem very quickly.
Guardrails therefore need to be treated as a strategic necessity, not a compliance footnote.
That means putting in place:
- Transparent workflows
- Clear review and approval points
- Explainability where decisions matter
- Strong data governance
- Release controls and QA processes
- Human oversight for high-impact outputs
The more embedded AI becomes, the more important these controls are. Speed without traceability is not innovation. It is exposure.
Modernisation is what makes AI scalable
From a technology perspective, one of the biggest mistakes businesses make is assuming AI can compensate for weak foundations.
It cannot.
AI amplifies whatever environment it enters. If the underlying systems are fragmented, the data is poor, the codebase is brittle, or engineering practices are weak, AI can accelerate the wrong outcomes. It creates more output, but not necessarily more value.
This is especially visible in software delivery.
AI can significantly improve engineering productivity. It can support coding, testing, review and iteration. But without modern architecture, small controlled change sets, disciplined review practices and shared technical understanding, it can also create larger, less understood risks at speed.
At Codurance, we often frame this simply: code is data. If that data is messy, AI creates more mess.
That is why modernisation and AI strategy belong together. Businesses that want real, repeatable value from AI need the right software foundations beneath it.
What does this mean for investors?
For private equity firms, AI is now part of both the upside case and the risk case.
It should influence how businesses are diligenced, how value creation plans are shaped, and how exit readiness is assessed.
That means looking beyond surface-level claims and asking harder questions:
- Is the business building on real proprietary data or knowledge?
- Is AI integrated into differentiated workflows, or sitting as a thin wrapper?
- Are the technical foundations strong enough to support scale?
- Does management understand where AI creates value and where it introduces risk?
- Are there meaningful governance controls in place?
- Does the business have the adaptability to evolve as the market shifts?
Traditional growth metrics still matter. But on their own, they are no longer enough to understand whether a business is positioned to defend and grow value in an AI-enabled market.
The Codurance perspective
At Codurance, we see AI creating the greatest impact when it is tied to strong software engineering, modernisation and disciplined execution.
That is why our work with investors and portfolio companies focuses on more than experimentation. We help organisations assess AI readiness, modernise legacy systems, build AI-enabled software responsibly, and establish the guardrails needed to scale with confidence.
Because in reality, AI is not a shortcut to value. It is a force multiplier.
And like any multiplier, it amplifies what is already there — good or bad.
The businesses that will create the most value in the years ahead are not the ones with the noisiest AI messaging. They are the ones with the clearest differentiation, the strongest foundations, and the discipline to turn AI into measurable commercial advantage.
Conclusion
The AI storm is real. But it is not a reason for panic. It is a reason for clarity.
For private equity firms and their portfolio companies, the path forward is not about chasing every new model or making sweeping automation claims. It is about understanding what is uniquely defensible, where AI can unlock practical value, and what controls are needed to protect the business as it evolves.
The winners will be those that double down on proprietary data and knowledge, stay close to customer needs, modernise their technology foundations, and apply AI with discipline rather than theatre.
