Outcomes from the panel discussion we hosted at the Association of the Luxembourg Fund Industry (ALFI) Private Assets Conference suggest that there’s a range of opportunities now within reach. These opportunities are focused primarily on operational challenges, but have further potential to create longer-term value.

However, all of our panelists cautioned that securing these benefits would require managers to grapple with issues such as how to manage data more effectively, how to update their existing technology stacks and how to overcome shortfalls in skills and knowledge.

Deriving value from the data

AI is already offering promising opportunities for growth, according to Vijay Sankaran, managing director, Artificial Intelligence at State Street. “Private markets is one of the most rapidly growing aspects of our business,” he said. “The challenge is that data in the private markets world is scattered and it needs to be harvested. We need to apply AI across these smaller fragments of information to look for large evolving patterns.”

Sankaran’s observation aligns with our recent research suggesting that alternative asset managers are excited about the potential to derive value from data — but they’re daunted by some of the practicalities of doing so. Almost three-quarters of the managers who took part in the research (73 percent) said they saw major opportunities in improved data management, including 10 percent that anticipated transformational benefits. However, managers said there were obstacles around internal collaboration, data integration and legacy technologies.

Overcoming these challenges can accelerate the use of AI and help managers gain greater efficiencies.

Promising operational improvements across the industry

Our panelists said there are several areas where the alternatives industry is already starting to deploy AI tools with the hope of seeing tangible results across operational domains.

“The corporate secretarial area is one where I see clear efficiency gains, in terms of our ability to manage minutes, investment committees, board meetings and so on,” said Cara Browne, head of Relationship Management and Provider Oversight at the Swedish investment organization EQT. “You do need an effective quality check — someone to check the transposition — but we’re very optimistic about this area.”

More broadly, Helge Baur, head of Private Markets Operations at global asset manager Allianz Capital Partners, which is part of Allianz Global Investors, said there’s potential to deploy generative AI on “everything that’s very text and language-intensive.”

That includes multiple use cases. To take just one example, Baur singled out the role of a fund controller: “When the controller comes into the office in the morning, one of the first things they need to do is to screen all the documentation received overnight,” he said. “They could just ask their AI assistant to do that job — ChatGPT, for instance, can tell them which are the most pressing topics.”

Nick Laird, founder and CEO of the technology company Verumex, said that even in areas of the alternatives sector such as real estate, where manual processes are still the norm, there’s now growing momentum for AI initiatives that synthesize managers’ data.

“Three or four years ago, a leading limited partner (LP) asked all the big funds to analyze their exposure to Amazon,” he said. “And in every case, it took between one and two weeks of people all over the world desperately trying to figure out the answer; today, we’ve built an AI-driven data product and we can do that work instantly. Straight away, you can see your exposure not only to Amazon but also to every sector and sub-sector. From a risk-management perspective, it’s pretty dramatic.”

Another area where AI has potential to give asset managers greater transparency is in their work to manage multiple service providers. “I’m receiving reports constantly,” said Browne. “Rather than having my service providers compile them manually, which is prone to error, I can see a real benefit in automating the data compilation and then using AI to interrogate it — that would be hugely powerful.”
 

AI’s potential to improve the investor experience

Looking further ahead, the panelists identified several areas where AI has the potential to deliver operational benefits, but is only just being explored. Client onboarding is a good example, where investors are often required to input the same information multiple times when they move money into new funds or start new relationships with managers.

“We need to make that a smoother process,” said Browne. “That would be powerful — we could onboard investors in such a clean and scalable way if we could figure this out.”

Another important priority identified by the panelists will be to support the demands of LPs for more granular data, particularly as they make portfolio decisions. Using AI to extract and interpret data, often in real time, could underpin a much deeper relationship with LPs that might once have depended on, for instance, a generic quarterly report and an occasional lunch meeting.

Panelists also believed there is potential to communicate more effectively with investors through the use of AI tools. “That’s one of the use cases that you can leverage technologies such as language-based models for,” said Sankaran. “We would like to be able to send out personalized messages to existing and potential investors using what we know about their investment and their profiles, their investment history, some of their preferences – in a customized way.”

The goal over the longer term should be to identify use cases for AI that build on the savings available from today’s deployments, said Baur – to focus on top-line benefits as well as the bottom line. “The first evolutionary step is to look at efficiency,” he said. “But that leads to a second step in terms of making money out of AI — can we onboard more investors, can we be quicker and, in the end, can we secure higher revenues?”

In the meantime, however, operational investment can create returns in other ways. For example, in our research, 58 percent of alternative asset managers surveyed said that a holistic data strategy was contributing to increased customer satisfaction.
 

How does AI measure up against human intelligence?

Though the global debate about AI’s impact on employment and labor markets continues, none of the panelists reported expecting technology to completely replace humans in the alternative assets space.

“AI is a support function — an assistant that delivers operational efficiencies by operating as an agent, rather than by taking a person out of the loop,” said Sankaran. “The reality is that when you put human experts in a room, you get differing points of view. That’s what makes the discussion special, and it’s not something AI will be able to do.”

However, AI is likely to create demand for new roles. One example flagged by the panelists is the growing need for “prompt engineers” — professionals who are highly skilled at working out exactly how to get the best out of AI engines. That won’t happen unless operators ask the right questions.

In the short term, the potential for AI is to help drive automation, allowing humans to engage in more value-additive work, according to Browne. “If we can shave some cost by getting AI to do things that makes us more efficient or scalable, that’s on our agenda,” she said. “But does that change the landscape here in Luxembourg for the skillset of the people we attract to work in finance? I don’t know the answer yet.”

Resolving these questions will be important as alternative asset managers think about the training they need to provide to existing staff, and where they will need to concentrate future recruitment efforts.

For now, our research identifies significant concerns about skills. More than four in 10 of the alternative asset managers surveyed (43 percent) said that one of the greatest challenges of getting more value from their data is existing staff lacking the time to stay on top of the latest technology developments and for training. A third (34 percent) picked out the difficulty of retaining skilled staff, and 33 percent said it was difficult to recruit staff with the right skills.

Resolving skills shortages is just one part of the challenge: Alternative asset managers are also trying to work out where to use resources to greatest effect — particularly as they integrate new tools into their existing technology stacks. According to our research, 80 percent of alternative asset managers said they believe they need to upgrade their existing technology by more than 25 percent.
 

Taking an incremental approach to implementing AI in the future

Although it’s tempting to think there will be a breakthrough moment for the use of AI in the alternatives sector, the reality is likely to be more prosaic. Our panelists encouraged managers to move forward iteratively, experimenting with new ideas instead of expecting a big-bang transformation to deliver everything they hope for all at once.

“It’s going to be incremental, because the truth is that this is hard and it’s expensive,” said Laird. “It’s not a case of inventing something new each day; this is learning as you go, learning what you don't know, and learning what's not working. That doesn't necessarily fit beautifully inside institutional environments where people want results.”

Moving cautiously will also be important because of the regulatory scrutiny of AI and broader concerns about the pace of technology. Baur referred, for instance, to the need to review materials produced by AI. “None of our communications to clients goes directly out of the machine — it’s all reviewed,” he said. “It’s the same across the whole of operations, there’s always a human in the loop.”

There are high expectations of AI and its benefits, but with an emphasis on a conservative approach. Alternative asset managers expect to tread carefully as they experiment with new technologies, and to focus on the most value-enhancing uses first.

“I would focus on the evolution of products in the alts space that will ultimately benefit the industry,” said Sankaran. “How we prepare the data, analytics and AI regulatory foundations for products that might become publicly traded and mainstream in 10 years from now – that’s the real question.”