Key Takeaways
- • Four in five US-based private equity firms have appointed a chief AI officer to oversee internal AI adoption, and two-thirds expect to allocate at least 25% of business budget to AI initiatives in 2026, versus fewer than half last year, according to an EY survey published at the end of 2025.
- • Global sponsors are moving AI into investment committee processes: EQT's Motherbrain has been elevated from deal-sourcing to IC sounding board, Advent's AIEdge is trained on 13 years of investment memos, and Warburg Pincus has embedded AI across daily workflows using 60 years of institutional knowledge.
- • Quick wins in internal adoption have concentrated in back- and middle-office functions and investor relations, while decision-support tools for due diligence and deal evaluation remain nascent and more costly, with Alvarez & Marsal warning that managers may spend tens of millions of dollars there.
- • Data quality, not AI capability, is the primary obstacle: WorkWise Solutions' Leigh Coney says pilot programmes most often fail because of poor data, and a China-based PE professional describes data scattered across personal laptops and knowledge trapped in individuals' heads.
- • Scale and nimbleness are contested as competitive advantages, with Advent (USD 94bn AUM) arguing its scale funds primary research while Japan-focused D Capital (about USD 700m AUM) argues its size allows faster AI adaptation.
Summary
Four in five US-based private equity firms have appointed a chief artificial intelligence officer to oversee internal adoption of the technology, according to an EY survey published at the end of 2025. Two-thirds of respondents expect to allocate at least 25% of their business budget to AI initiatives in 2026, compared with less than half last year. Accompanying commentary argued that the debate over adoption is settled and that investors must now show how they use AI to "create differentiation and not just 'get to average.'" Karen Sands, COO for private equity at Federated Hermes, said sustainable advantage comes from deploying AI "in a controlled and measurable way" that delivers tangible business value while maintaining risk management and human oversight.
Global sponsors are leading the way, with AI increasingly present at investment committee level without replacing human decision makers. EQT's Motherbrain engine has moved from deal-sourcing to acting as an IC sounding board. Advent uses AIEdge, trained on 13 years of investment memos including deals it passed on, to challenge assumptions in every opportunity presented to the IC. Warburg Pincus says it has embedded AI across daily workflows, using an engine that analyses decades of investment memos, board materials and presentations. Jesse Thomas, head of the global data science team at Advent, argued that scale still matters, since the firm can fund more surveys, expert calls and advisor networks than smaller GPs, and said AI does not close that gap.
Middle-market managers counter that nimbleness matters more than scale. D Capital, a Japan-focused firm with around USD 700m in AUM, believes commercially available IT systems are sufficient for most purposes and runs AI adoption as bottom-up experimentation under top-down governance. Megumi Matsutani, partner and chief data officer at D Capital, said smaller size lets the firm adapt quickly, while large funds must manage risk globally and localise across regions. Internal adoption typically starts with process-oriented automation such as summarising information, project management and drafting documents. At Federated Hermes, the quick wins have come in back and middle offices and in investor relations. Deal-making use cases such as due diligence and investment memo drafting are harder, and Daniel Angelucci, managing director at Alvarez & Marsal's performance improvement practice, warned that substantial investment in agentic AI and automation will be needed over the next decade just to maintain baseline competitiveness.
Advent's dedicated AI agents produce initial reports from company details and industry comparables, help analysts compile deal documents, run daily searches for updated information and check presentations and spreadsheets before they reach senior partners. Thomas said the first 70% of due diligence is largely common across a sector and is produced against a proprietary codebase to guard against hallucinations, while the remaining 30% is bespoke work. Much adoption is grassroots: employees at smaller managers use licences for ChatGPT, Claude and Gemini under data-privacy guidelines. Affirma Capital's India-based investment professional built a pipeline tool now in beta testing.
Data is widely seen as the binding constraint. An investment professional at a China-based PE firm said data is not standardised and that important knowledge is trapped in individuals' heads. Leigh Coney of WorkWise Solutions said AI pilots most often fail because of poor data quality and that larger players are still implementing Salesforce. BlackRock, through its private equity team and BlackRock Systematic, says it has spent more than a decade developing proprietary predictive models; Johnathan Seeg, global co-head of BlackRock Private Equity Partners, said the growth of digital data around private companies has changed private investing. Jolt Capital's Jolt Ninja platform, which the article says launched in 2016, draws on five million companies and 10 million patents to benchmark regional valuations. According to managing partner Jean Schmitt, it flagged better risk-return in Japan and South Korea than the US and supports a 25-company shortlist in Japan without a full local team. Bain Capital has asked teams to redesign workflows around an "AI twin" of existing processes, according to Jordi Diaz, head of data science and AI at the firm.
Global sponsors are leading the way, with AI increasingly present at investment committee level without replacing human decision makers. EQT's Motherbrain engine has moved from deal-sourcing to acting as an IC sounding board. Advent uses AIEdge, trained on 13 years of investment memos including deals it passed on, to challenge assumptions in every opportunity presented to the IC. Warburg Pincus says it has embedded AI across daily workflows, using an engine that analyses decades of investment memos, board materials and presentations. Jesse Thomas, head of the global data science team at Advent, argued that scale still matters, since the firm can fund more surveys, expert calls and advisor networks than smaller GPs, and said AI does not close that gap.
Middle-market managers counter that nimbleness matters more than scale. D Capital, a Japan-focused firm with around USD 700m in AUM, believes commercially available IT systems are sufficient for most purposes and runs AI adoption as bottom-up experimentation under top-down governance. Megumi Matsutani, partner and chief data officer at D Capital, said smaller size lets the firm adapt quickly, while large funds must manage risk globally and localise across regions. Internal adoption typically starts with process-oriented automation such as summarising information, project management and drafting documents. At Federated Hermes, the quick wins have come in back and middle offices and in investor relations. Deal-making use cases such as due diligence and investment memo drafting are harder, and Daniel Angelucci, managing director at Alvarez & Marsal's performance improvement practice, warned that substantial investment in agentic AI and automation will be needed over the next decade just to maintain baseline competitiveness.
Advent's dedicated AI agents produce initial reports from company details and industry comparables, help analysts compile deal documents, run daily searches for updated information and check presentations and spreadsheets before they reach senior partners. Thomas said the first 70% of due diligence is largely common across a sector and is produced against a proprietary codebase to guard against hallucinations, while the remaining 30% is bespoke work. Much adoption is grassroots: employees at smaller managers use licences for ChatGPT, Claude and Gemini under data-privacy guidelines. Affirma Capital's India-based investment professional built a pipeline tool now in beta testing.
Data is widely seen as the binding constraint. An investment professional at a China-based PE firm said data is not standardised and that important knowledge is trapped in individuals' heads. Leigh Coney of WorkWise Solutions said AI pilots most often fail because of poor data quality and that larger players are still implementing Salesforce. BlackRock, through its private equity team and BlackRock Systematic, says it has spent more than a decade developing proprietary predictive models; Johnathan Seeg, global co-head of BlackRock Private Equity Partners, said the growth of digital data around private companies has changed private investing. Jolt Capital's Jolt Ninja platform, which the article says launched in 2016, draws on five million companies and 10 million patents to benchmark regional valuations. According to managing partner Jean Schmitt, it flagged better risk-return in Japan and South Korea than the US and supports a 25-company shortlist in Japan without a full local team. Bain Capital has asked teams to redesign workflows around an "AI twin" of existing processes, according to Jordi Diaz, head of data science and AI at the firm.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.