AI is everywhere - in pitch decks, IC meetings, boardrooms, and break rooms. Leaders and line staff in every sector are aware that they need to do something, though many are unsure what. But amidst the hype and handwringing, the private equity world is busy getting on with it.

Unlike many industries still stuck in exploration mode, PE firms – entrepreneurial by nature and data-hungry by design – are already experimenting and executing, recognizing AI’s disruptive power and potential to deliver strategic advantage. We spoke with some senior professionals to understand what’s real, what’s next, and what’s noise. Is AI really the ultimate paradigm shift – at least, until the next one?

As AI has become increasingly mainstream, the conversation has shifted from whether it will impact the world of alternative investment to how and where it can best be harnessed to deliver value.

In talking to people around the industry, it’s clear that many PE firms are already a long way down the road on their AI journeys. Some, including one of our interviewees, have established cross-functional AI committees to lead the exploration of AI’s potential and to direct investment. Some have developed or are developing their own in-house AI tools, while others are focusing on third-party tools. But while the build or buy dilemma may be a familiar one, it’s important to note that no-one is looking at building their own LLM. The major AI players have planted their flag in that territory and, for the foreseeable future, PEs and others are focusing more on tools, or more precisely ‘bots’ that they can layer onto these engines to deliver productivity gains and novel insights. These tools are likely to be very specifically focused, built to conduct very narrow, very deep functions.

Perhaps unsurprisingly given the nature of industry, some have addressed the issue by investing in software businesses with the expectation of using them to develop AI tools using capabilities that go beyond their own in-house teams.

Which AI's are PE's using?

Asked which AI they are using, our contacts listed both generic and industry-specific tools:

Putting AI to work

PEs are already using AI extensively and across a number of important task areas, some specific to PE and others reflecting the general business uptake of AI, particularly LLMs, to accelerate day-to-day tasks.

Our PE interviewees, for example, mentioned analysts and others using AIs to draft investment memos and summarize research reports –much like everybody else does.

But, more specific to the sector, PEs see clear productivity benefits in deploying AI in the origination stage of the investment lifecycle: ‘There are so many companies. AI definitely has a role to play in finding those unseen potential fits,’ said one PE.

What is unclear at the moment is whether PEs are using AIs to conduct general prospecting or whether they are trying to teach the AI to manifest their ‘secret sauce’ in their searches. The frontier probably lies in merging these two approaches.

One clear theme that we are seeing among PEs is concern about relinquishing too much control and decision-making to AIs. Most recognize AI’s power to weed out misfits using relatively simple rules, while ensuring that few or no prospects are missed. But, as the funnel narrows and the focus switches to risk, PEs currently seem unconvinced that AI can adequately interpret and apply ‘human’ interpretations of risk. In fact, precisely where ‘human’ intervention is important – whether that means an understanding of risk, fidelity to a ‘secret sauce’ or ‘intuition’ based on experience – seems likely to be most hotly contested issue in PE.

Beyond origination

PEs are also using AI to analyze online data rooms and extract in seconds the kind of nuanced insights that would take analysts hours to find. For example, for any given company, an AI can be tasked with identifying its top 10 customers – who and where are they; how secure do the relationships look; are the products/services they buy under pressure? Who are its competitors and what pressures are they facing?

PEs are then doubling up, using AIs to draft investment memos based on these AI-derived insights, and to generate bespoke emails to prospects. AIs are increasingly used by analysts to build an understanding of the commercial landscape in which companies operate - the pressures they face and the social sentiment about them. ‘With AI,’ said one interviewee, ‘we can ask questions we simply couldn’t have asked previously.’ It’s also clear that some PEs are doing this alongside their conventional research and then comparing the outputs.

‘It’s at the top of the funnel where it will be most useful.’

Beyond the origination stage, PEs see a more limited role for PE: ‘It’s at the top of the funnel where it will be most useful,’ said one contact. ‘Further down, I expect the professional consultants we use today will still play a significant role.’

Will AI change the way due diligence is conducted? Perhaps, said our interviewees, but only in a limited way at this stage. ‘PEs have big budgets for this kind of thing,’ said one interviewee. Due diligence is more about knowing the right questions to ask and then marshalling resources. There are a lot of experts to provide the insights we need.’ However, there are some specific use cases where AI has further potential. In a reprise of its role as an origination tool and reflecting current trends within the alternative investment ecosystem, our interviewees highlighted the potential to use AI to identify potential bolt-ons or new customers for prospects and portcos. In these use cases, there is also a frontier to breach. While it may be perfectly satisfactory to challenge an AI to look around, up and downstream for bolt-ons, the real value is likely to come when PEs find a way to set AI more leftfield challenges. As Sam Altman, founder of OpenAI, suggested, to get the best results – and, if we’re honest, to train it – ‘feed it the toughest business problem.’ The challenge in this instance will be to develop the skills to frame and refine the right kinds of questions.

Our interviewees also alluded to cases where AIs had been used to identify the top consultants in any given sector/geography, either to support due diligence activities or to speak at profile-building events. It’s easy to see a similar approach being taken to identify potential sponsors for events.

At the value creation stage, PEs see real challenges for AI to make any significant impact. Chief among these is the patchwork nature of information within a portco - or a whole portfolio. ‘It’s hard, if not impossible,’ said one interviewee, ‘for AI to understand the assumptions and factors that drove decisions or to see how a company is performing against the investment case - it’s stymied at every turn.’

The verdict is clear: AI is an accelerator, not a replacement. But, that is today – and may be more a function of the limits of the users’ faith than of the AI itself.

Major gains

PEs are already realizing huge productivity gains using AI to search both for companies and then for company information. It’s great for what are described as ‘quick and dirty’ searches - who are the major players, how big is the national/regional market, how fast is it growing?

‘Obviously there are hallucinations, so it’s important you don’t take the output as gospel, but there are huge savings in time to be made in tackling manual, volume and repeat tasks,’ said one contact. While they acknowledge that some errors will slip through, the gains that AI offers in terms of speed to build a broad understanding of a market or

opportunity far outweigh the requirement to then go back and fact check the information it provides. As one of our interviewees put it: ‘Ultimately, [AI] is a tool, right? And anything that looks off needs checking.’

And, it seems, AI couldn’t have come at a better time for the PE sector, at least according to our contacts. The world, they say is becoming more confusing - the sector is more crowded, the processes more sophisticated and the ‘easy pickings’ that characterized the early stages of PE are long gone. The focus has pivoted away from cost reduction to growth as a means of creating value, and information, data and knowledge have become much more important. As one interviewee put it: ‘Decision making has to become more data-based, and less testosterone-driven.’

‘Decision making has to become more data-based, and less testosterone-driven.’

AI reshaping the alternative investment landscape

AI is already changing the nature of search and origination. Currently, it is not uncommon for a PE firm to sign multiple NDAs to do one deal. AI will enable PEs to conduct more-targeted, more-textured searches, for example, to identify companies that are likely to encounter succession challenges: ‘Show me companies whose founders are over 70 years old and don’t have a succession plan.’ That kind of focus will greatly streamline the origination and relationship-building phases of the lifecycle.

AI also looks set to reshape the structure of PEs themselves, and quickly. ‘PEs will need fewer analysts to conduct the volume work at the mouth of the funnel,’ said one interviewee. ‘But at the top end, the relationships and personal approach will remain crucial. Ultimately, it will still take humans to negotiate terms and go back and forth – I think that will always be hard to do through technology.’ And, perhaps paradoxically, AI seems set to enhance the value of data itself– or at least proprietary data. Just as AI is making publicly available data accessible in an instant, it will drive up the value and utility of data and analytics that are not publicly available. As one contact put it: ‘There are a lot of models out there, but if you have better or different data than everyone else, if your AI model is 5% better because of that, especially in finance, you’re at a massive advantage.’

Potential downsides

Alongside all the positives, the industry also recognizes some threats. The growth in reliance on AI will, it is thought, inevitably lead to data breaches and vulnerability to cyberattacks. In fact, in the case of the latter, AI may even be the agent driving the attacks. ‘It’s an asymmetry in terms of information and expertise,’ explained one of interviewees. ‘Some people will be less expert in AI than others and will, as a result, be vulnerable. I see AI being used by criminals to scam at scale, targeting people and businesses that don’t know AI as well.’

More prosaically, reliance on AI in PE, and the adjustment in behavior that will go with it, will mean that mistakes will get through: ‘If AI is right 99% of the time, it’s easy to see how people will lower their guard. It’s human nature. So, yes, mistakes will get through.’

‘It’s a tool isn’t it? And those firms that use the tool effectively will displace those who don’t.’

Shifting that paradigm

For the time being, people with deep investment experience seem reluctant to cede decision control to AI, partly out of a lack of faith and partly, we suspect, out of a degree of vanity. There will be job losses and role obsolescence around the wider alternative investment ecosystem but how quickly that effect moves up the line remains to be seen.

As AI develops, we believe it will start to yield real advantages in some of the so-called ‘edge cases’ that are currently the preserve of experienced investment brains. Across valuation, value bridge, portco management and even LP relationships – areas where context, perspective and nuance rule – AI will start to become a powerful aide and even a partner, learning from its users’ behaviours and predicting their likely responses in any given set of circumstances.

Ultimately, the PEs we spoke to see AI as an enormous opportunity to enhance what they already do rather than a threat to their existence: ‘It’s a tool isn’t it? And those firms that use the tool effectively will displace those who don’t.’ Whether you like the phrase ‘paradigm shift’ or not, AI is a part of alternative investment now, and its role is only going to grow.