The Promise and Reality of AI in Retail Investing

For a growing number of retail investors, AI chatbots have become a first stop for financial guidance – always available, endlessly patient, and seemingly all-knowing. Tools like ChatGPT can quickly explain ETFs, the relationship between risk and return, or the basics of long-term wealth building. According to consumer protection experts, that kind of basic educational role is where AI genuinely helps.

Yet behind the surface convenience, a much more sobering picture emerges. A review by Germany’s Stiftung Warentest found that more than a third of AI-driven investment funds had been liquidated since the group began tracking them, and the surviving funds showed no consistent edge over traditional benchmarks. Niels Nauhauser of the consumer advice centre in Baden-Württemberg warns that AI-generated answers routinely contain factual errors, oversimplifications, and even fabricated sources – problems that are impossible to spot without substantial prior knowledge.

The technology is already embedded in many corners of finance: banks use AI to flag suspicious transactions and assess creditworthiness, while robo-advisors build and monitor automated portfolios based on a client’s risk profile. Industry lobby Bitkom points out that these applications operate under a dense web of financial supervision, data protection rules, and the EU’s AI Act. But experts caution that regulation alone cannot fix the core problem: AI models reflect the biases in their training data, often struggle to distinguish between fact and opinion, and can reverse their recommendations entirely after a slightly rephrased question.

Why AI-Powered Investment Advice Often Fails

AI Funds Have Delivered Little to No Edge

Stiftung Warentest’s tracking of funds that explicitly use AI for stock selection paints a disappointing picture. Over one third have already been shut down, and the remainder show mixed results. “Some beat their comparison index, some don’t,” says the organization’s investment expert Yann Stoffel. “In short, we currently see no advantage in AI funds.” This suggests that the supposed data-processing superiority of AI has not yet translated into reliable outperformance for ordinary investors.

The Transparency and Liability Gap

A deeper structural issue is transparency. Because users cannot see how an AI arrives at a specific recommendation, they cannot critically assess its reasoning. Compounding the problem, some AI providers have commercial partnerships that can skew the information they surface. “Such partnerships can lead to certain content, sources, or viewpoints being favoured,” warns Nauhauser. “For consumers, it’s not transparent what interests are at play in the background.”

Unlike a regulated bank advisor or even a robo-advisor, a general-purpose AI chatbot assumes no legal responsibility for its output. “The AI isn’t liable for its tips,” Stoffel notes. A human advisor or a licensed digital service, by contrast, can be held to account for documented misadvice.

Regulation Exists but Doesn’t Fix the Core Problem

The financial sector’s use of AI is subject to oversight by banking supervisors, data protection authorities, and the EU’s AI regulations. That provides a safety net for structured products like robo-advisors. However, generic AI assistants that offer impromptu investment commentary fall into a grey area where the same regulatory rigor does not apply. The result is a marketplace where retail investors can easily get plausible-sounding but ultimately unreliable guidance with no recourse.

What Individual Investors Should Do Now

  • Use AI only for learning, not for trading decisions. Ask it to explain what an ETF is or how compound interest works, but never for specific buy-or-sell recommendations.
  • Verify every answer with independent sources. Before acting on any AI-generated insight, check the same topic on trusted, non-commercial sources such as consumer advice sites or financial regulator publications.
  • Be sceptical of free tools and their hidden biases. If an AI consistently mentions certain products or providers, ask whether a business partnership might be behind it – such relationships are rarely disclosed.
  • Compare robo-advisor track records carefully. While regulated robo-advisors can lower advisory costs, their performance still depends on the underlying algorithms; look for multi-year, verified track records, not back-tested promises.
  • Build your own financial literacy first. Experts agree that only a solid foundation of knowledge lets you spot the errors, biases, and logical gaps in AI-generated advice.