TALKING POINT

oftware Improvement Group( SIG) has published AI

Ambition and Reality in Private Equity – new report examining why expectations for artificial intelligence in private equity increasingly outstrip operational reality, and how that gap is beginning to influence deal risk value creation and exit outcomes.

This disconnect is emerging just as private equity faces expensive assets swollen exit backlogs slow liquidity and buyout returns that increasingly look modest compared with public markets. In that environment investors need clearer visibility into which AI moats are genuine and which remain largely narrative.

AI has rapidly shifted from experimentation to a central pillar of deal theses value-creation strategies and portfolio narratives. While 88 % of organisations report using AI and 64 % say it enables innovation deployment across portfolio companies remains uneven and governance is often weak once systems move into production.

Within private equity specifically roughly two-thirds of general partners report AI pilots somewhere in their portfolios. Yet only about 40 % say AI is embedded across multiple business processes revealing a widening gap between ambition and operational reality.

• AI is present but far from dominant. Of all production systems analysed by SIG in 2025 only about 1.5 % qualified as AI systems indicating that most portfolios remain early in enterprise-scale adoption.

• Quality is a material concern. Around 72 % of AI systems scored below SIG’ s recommended build-quality threshold raising risks related to maintainability security and regulatory compliance.

• AI-assisted coding delivers mixed results. Reported productivity outcomes range from a 19 % slowdown to a 26 % speed increase depending on context and controls. In SIG experiments AI-generated code also showed roughly double the security-risk violations compared with similar humanwritten projects.

• Security risks are amplified rather than replaced. AI introduces new attack surfaces through data pipelines models prompts and external dependencies adding exposure on top of existing cyber risk.

• Governance gaps widen as regulation fragments

• Regulatory divergence now complicates diligence reporting expectations and cross-border portfolio governance. •

“ AI is changing the way we do business fundamentally and reliance on technology is increasing to previously unseen levels,” said Luc Brandts, CEO, Software Improvement Group.“ Yet there is still no clear oversight. This is not a new problem but AI is increasing the stakes tenfold. Can we trust what is being built? Are we investing in the right places? How ready are we for AI? These are questions every investor must answer.”

AI ambition meets investment reality

The report finds that private equity firms are not struggling with AI primarily because of the technology itself. Instead, the challenge lies in the underlying software foundations inside portfolio companies which often remain opaque and poorly documented. These foundations frequently cannot support AI at scale. As a result, AI maturity is increasingly influencing how investors price risk at entry execute during the hold period and defend value at exit.

Key findings include:

• The technical moat is under pressure. As AI accelerates the creation of generic functionality defensibility increasingly depends on assets that are difficult to copy domain expertise proprietary data scalable architecture and disciplined engineering practices.

18

INTELLIGENT CIO NORTH AMERICA www. intelligentcio. com