As AI transforms financial services, the alternative investment industry is entering a new phase of transformation. In this interview, Joachim Meyer, Senior Manager at PwC shares his perspective on the opportunities, challenges, and long-term impact of generative AI in alternative investments, as well as Luxembourg’s positioning in this evolving landscape.

Artificial intelligence and generative AI are rapidly transforming the financial sector. In your view, what are the main changes currently at play in alternative investments, and where do you already see high-value use cases?

In alternative investments, generative AI represents a major opportunity. Historically, the industry has been highly manual and relationship-driven, operating through high-touch interactions, extensive documentation, and close collaboration between investors and investment firms.

AI is not replacing that human dimension, but it is transforming the operational layer surrounding it. Today, firms still spend enormous amounts of time processing partnership agreements, due diligence questionnaires or quarterly reporting packages. Many of these workflows remain manual and fragmented across teams.

This is where AI already delivers value. One key use case is document intelligence: extracting, structuring, and analysing information from large volumes of unstructured content. Another high-value area is improving data transparency and explainability around financial figures.

Alternative investments have specific characteristics compared to traditional markets. What constraints does this create for the use of AI, and how can they be addressed?

Alternative investments are very different from traditional listed markets. A large part of the information is private, bespoke, negotiated, and not standardised. This creates specific constraints for AI.

The first constraint is fragmentation. Data is often spread across fund administrators, AIFMs, investment managers, legal advisors, depositaries, internal teams, external portals, emails, Excel files, and PDFs. Before AI can create value, firms need to know which information is available, where it sits, and whether it is reliable.

The second challenge is the lack of standardisation. If you compare quarterly reports, capital account statements, side letters, or due diligence questionnaires from different managers, formats can vary significantly. Unlike traditional markets, where data is often available in structured feeds, much of the relevant information in private markets is embedded in documents.

The third challenge is data reliability. In Alternatives, the issue is often not only access to data, but also confidence in the data. Is this the latest validated version? Has it been reviewed? Is it consistent with the administrator, the fund documents, the investor register, or the financial statements?

The way forward is not to deploy AI in isolation. What works best is a combination of AI, data governance, workflow, and human oversight. AI can collect, extract, classify, and analyse information, but humans remain essential to validate, certify, and approve the data before it is reused for reporting, decision-making, or investor communication.

To what extent can AI truly improve decision-making for fund managers and where do its current limitations lie?

AI improves decision-making by enabling firms to analyse large volumes of documents and extract insights that would otherwise require significant manual work. It is also very useful for generating first drafts of due diligence questionnaires, investment memos, or reports, allowing teams to focus on review rather than starting from scratch.

Another key benefit is consistency. Reviewing repetitive documents manually can become mechanical over time, whereas AI can apply the same logic across large volumes of information in a far more systematic way.

However, there are still important limitations. Hallucinations remain a risk, especially in finance, where AI can generate convincing but inaccurate information. Human validation therefore remains essential. AI also depends heavily on the quality of the information it is given. If the underlying data is incomplete, outdated, inconsistent, or poorly governed, the output will not be reliable.

Today, AI is seen as a very fast junior analyst: highly efficient at reading, comparing, and summarising information, but still requiring human supervision and final approval.

In a sector where human judgment remains central, how do you see the role of experts evolving alongside AI? Can we speak of a shift in posture?

There is clearly a shift in posture. Experts are moving from being producers of content to becoming reviewers, challengers, and orchestrators. The value no longer lies only in knowing the answer, but increasingly in being able to assess whether the AI’s answer is actually correct and explain why.

Human judgment therefore becomes even more important. AI can generate analysis very quickly, but experts still need to validate outputs, challenge assumptions, and make decisions in situations where there is limited or imperfect data.

At the same time, alternative investments remain a relationship-driven industry. Investors trust people, not algorithms. Credibility, experience, communication, and the ability to explain complex topics clearly will remain critical.

For me, AI is similar to autopilot in a plane: it can assist and automate many tasks, but you still need the pilot understanding and supervising the system and making the final decisions.

Luxembourg is a major hub for investment funds. How can the country leverage AI to strengthen its position within the alternative investments ecosystem?

Luxembourg already has a major advantage: the density of its ecosystem. The country brings together alternative investments fund managers, administrators, legal firms, and service providers within a highly international and multilingual environment.

The next step is to become more than simply a fund domicile. It has the opportunity to position itself as a centre of excellence for AI. Many of the operational processes that can benefit from AI are already concentrated in Luxembourg: fund administration, investor services, reporting, compliance, legal documentation, oversight, middle-office, back-office, and data management. These are not theoretical use cases. They are real operational pain points where AI can be deployed at scale.

Luxembourg also has an important role to play around trust, regulation, data governance, and operational resilience. In financial services, AI adoption will not only depend on the sophistication of the models, but also on the ability to control data, document decisions, manage risks, and comply with regulatory expectations.

This is where Luxembourg can differentiate itself: by combining fund expertise, multilingual talent, regulatory credibility, data sovereignty, and practical implementation capabilities. The ambition should not simply be to experiment with AI, but to industrialise responsible AI in real fund operations.

What do you see as the main challenges? And looking ahead, what could alternative asset management look like in 5 to 10 years with the rise of AI?

The main challenges are not only technological. They are also regulatory, organisational and cultural. In financial services, AI must be deployed within a controlled environment, with clear governance, data protection, cybersecurity, operational resilience and human oversight. Frameworks such as the EU AI Act, DORA, GDPR and sector-specific regulations such as AIFMD - and, where relevant, MiFID - increase expectations around transparency, accountability, risk management and control. This is particularly important in alternative investments, where firms handle confidential investor information, complex fund documentation and operational processes involving multiple parties.

A second challenge is the quality of the foundations. AI is only as reliable as the data, processes and controls around it. If a firm does not know where its data comes from, whether it is complete, who validated it, or whether it is the latest approved version, AI can amplify existing weaknesses rather than solve them. This is why successful adoption will depend not only on the choice of AI tools, but also on strong data architecture, governance, validation workflows, auditability and clear accountability between technology teams, operations, compliance and business experts.

Over the next five to ten years, AI will become embedded in the operating model of alternative asset managers and service providers. It will help teams search fund documents, draft DDQ responses, review reporting packs, detect exceptions, support compliance checks, reconcile data and explain movements more efficiently. The main impact will be productivity, scalability and consistency: allowing firms to grow assets, investors, products and reporting obligations without increasing headcount at the same pace. But the real differentiator will not be who has access to AI; most firms will. The differentiator will be who has the right foundations, controls and expertise to use it safely, effectively and at scale.

For more information visit the webpages: Process Data Bridge and How Process Data Bridge transformed Private Equity fund operations.