There’s a good chance your pension fund already uses AI somewhere, and the platform your ISA sits on too. What matters is whether the people managing your money understand how it’s used, and whether they can explain it to you plainly.

This piece is for you as a client, not for the advisers and firm owners I usually write for. If you’re going into a review meeting and want to ask better questions about how technology shapes decisions about your money, it should help.

Where AI is used in investment management

AI shows up at several different points in managing investments, and what it means for you depends on where.

Portfolio construction and rebalancing. Many discretionary fund managers use AI-assisted tools to scan large numbers of holdings, spot drift from target allocations and generate rebalancing instructions. The fund manager still makes the final call, but more of the first analysis is automated.

Risk monitoring. AI systems can check a portfolio against a defined risk profile all the time, flagging when a holding or asset class moves outside agreed limits. A human analyst can’t watch hundreds of positions at once, so this is useful. The AI flags, and a person decides what to do.

Market data analysis. Quantitative strategies have used algorithms for decades. What’s changed is scale and cost: models that once needed a dedicated quant team are now available to much smaller fund managers through third-party platforms.

Regulatory reporting and compliance. You won’t see this part, but it matters. Under MiFID II and related rules, firms have to show suitability, record transactions accurately and produce structured reports. AI tools help assemble and check that paperwork, which reduces manual errors.

What you shouldn’t find at any reputable firm is an AI system making investment decisions in a regulated setting with nobody overseeing it. That isn’t how the industry works today, and a claim otherwise deserves a careful look.

The benefits for your portfolio

The real benefits are mostly about consistency, speed and coverage, not about the AI being cleverer than a human analyst.

A rules-based system applies the same criteria every time. It doesn’t have a bad day, forget a criterion or skip a step because it’s busy, which is valuable for compliance and risk checks. It can also check a portfolio against several risk limits in seconds, which matters most in fast-moving markets where spotting an out-of-tolerance position late can be expensive.

Then there’s volume. One analyst can only read so much. AI tools can take in market data, company results and economic indicators at a scale no human team can match.

The honest case for AI in investment management is that it gives human judgement better and faster inputs to work with, rather than replacing it.

All of that depends on the quality of the system, the quality of the data going in, and the quality of the human oversight on top.

The risks for your portfolio

Your adviser should be able to talk through the risks as plainly as the benefits.

Reliance on historical data. Most AI models learn from the past. A system trained on the last ten years of markets hasn’t seen every kind of shock history can produce. That’s no reason to avoid AI tools, but it is a reason to treat an AI-generated risk estimate as an estimate, not a precise forecast.

Opacity. Some AI systems, especially those built on deep learning, can’t easily explain why they made a recommendation. Regulated advisers have to be able to show the reasoning behind a suitability decision, so if your adviser uses a tool they can’t explain to you, it’s fair to ask about it.

Concentration. Much of the industry’s AI runs on platforms that depend on a small number of underlying AI vendors. If a widely used system makes a systematic error, it can spread across many firms at once. EIOPA’s AI governance principles for insurers and pension funds address this kind of risk[1].

Governance gaps. The FCA and Bank of England’s joint statement in May 2026 signalled that AI governance for UK-regulated firms is moving from voluntary good practice towards a formal expectation[2]. A firm that can’t show how it supervises its AI tools, checks their output and holds someone accountable is taking on regulatory risk, and your money sits in the same system.

What it means for you as a client

Your suitability profile matters more when AI tools are in use, not less. If the system watching for portfolio drift checks you against your stated risk profile, that profile has to be accurate and current. An out-of-date fact-find means the AI is aiming at the wrong target.

Accountability stays with people. The AI tools your adviser or fund manager uses feed into decisions; they aren’t decisions. If something goes wrong, the regulated firm is responsible, not the software, and it’s reasonable to expect them to explain how they supervise the tools they use.

How your data is handled is a fair question too. Your personal and financial information may be processed by the AI systems your adviser uses. Reputable firms will have clear policies. If you’re not sure, ask.

Four questions for your next review meeting

You don’t need to be an AI expert to ask good questions. These four are worth putting to your adviser directly.

Do you use AI in managing my portfolio? “Do you use any AI or automated systems in how you manage client portfolios, and can you explain what they do?” A good adviser will have a plain answer. If they can’t explain it, note that.

Who checks what it flags? “When an AI system flags something about my portfolio, who reviews it and what happens next?” The answer should make clear a person is involved in any material decision. If it’s fully automated with no human review, ask more.

Is my profile up to date? “Are my risk profile and circumstances up to date, and how often are they reviewed?” Always a good question, and more so when automated monitoring is involved.

What happens when it’s wrong? “If an automated system made an error that affected my portfolio, how would you find it and what would happen?” This tells you whether the firm has thought about failures, not only successes.

What a firm using AI well looks like

A firm that uses AI well can tell you what it does, what it doesn’t do, and where a person is involved in every decision that matters. It treats its AI tools like any other outside supplier, with documented oversight, clear accountability and a plan for when things go wrong.

The firms I’d worry about present AI as a black box that produces reliable answers, can’t explain the reasoning behind a recommendation, or use automation in place of professional judgement rather than as an input to it.

You don’t need to audit your adviser’s technology. You do deserve a plain explanation of how the tools they use serve your interests, and the four questions above are a good place to begin at your next review.

This article is for informational purposes only and does not constitute regulated financial advice or a compliance opinion. Consult a qualified compliance professional for advice specific to your firm.

Sources

  • [1] EIOPA, Artificial Intelligence Governance Principles, Towards Ethical and Trustworthy AI in the European Insurance and Occupational Pensions Sector, published 2024 (referencing end-2024 AUM figures). Risk-management and investment-strategy context.
  • [2] FCA and Bank of England, joint statement on AI governance for UK-regulated firms, May 2026. Signals shift from voluntary best practice to formal supervisory expectation.