Why Investors Still Do Not Fully Trust AI Stock Advice

Artificial intelligence (AI) adoption continues to accelerate at “historic speed.” As the technology keeps improving and reaches more people than ever, it is now quietly becoming part of the retail investor’s toolkit as well.

From screening stocks to summarizing earnings calls, creating trade ideas, analyzing portfolios, and executing trades, AI is fast becoming an integral part of investing.

But despite impressive advances in generative AI and machine learning, there remains a reluctance among investors to rely on AI for actual buy-and-sell decisions. Investors are happy to ask an AI model to flag a risk or explain a balance sheet, but fewer are willing to trust the same model when real money is at stake.

This shows that trust is lagging behind capability, with investors concerned about hallucinations, algorithmic bias, data quality, cybersecurity, and AI’s inability to fully account for unpredictable macroeconomic or geopolitical events.

As the industry and regulators work out what accountability looks like, most investors are turning to AI as a tool to support their decision-making rather than a replacement for human judgment. The question that remains is: will retail investors trust AI enough to let it influence portfolio decisions?

AI Is Reshaping the Investment Playbook

AI is experiencing a booming market, with the technology expected to drive significant economic growth over the coming decades. Compared to railroads, electric power, and the internet, the transformative technology promises to upend everything from everyday internet searches to deep industrial workflows, complex manufacturing, and drug discovery.

The rapid advancement of AI is also influencing financial markets. Today, financial institutions, brokerages, hedge funds, and fintech firms are employing AI across their platforms for a wide range of tasks.

These entities are using AI for sentiment analysis, risk modeling, fraud detection, quantitative trading, portfolio optimization, personalized investment recommendations, and customer support.

Generative AI has further expanded these capabilities, offering significant advantages for the regular investor. It allows investors to query complex financial information using natural language, lowering the barriers to sophisticated market research.

Then there’s AI’s ability to analyze enormous amounts of data continuously and at a much faster speed than a human analyst, who could take days to work through it all.

While analyzing volumes of structured and unstructured data, the technology monitors thousands of assets simultaneously and identifies hidden patterns. It also automates repetitive research tasks and reacts almost instantly to new information.

Being a machine, AI isn’t driven by fear or greed either; rather, it applies consistent methodologies that reduce emotional biases like overconfidence, loss aversion, and herd-following that distort human decision-making.

All these capabilities make investment research more efficient and accessible, especially for retail investors who previously lacked institutional-grade analytical tools.

On top of all this, AI has been found to outperform humans in stock picking, beating “93% of managers over a 30-year period by an average of 600%.”

AI, however, isn’t infallible. There are significant limitations to this technology, including the fact that most advanced AI systems operate as black boxes, meaning there’s no way of knowing how they generate their recommendations.

The technology is also only as good as the data it was trained on. As a result, AI models can produce not just inaccurate but fabricated information.

The hallucination problem with AI is well-documented. Not only does it give incorrect answers, but it delivers them with confidence, making them harder to detect.

“Hallucinations are plausible but false statements generated by language models. They can show up in surprising ways, even for seemingly straightforward questions,” explained OpenAI in its blog titled ‘Why language models hallucinate.’ “For example, when we asked a widely used chatbot for the title of the PhD dissertation by Adam Tauman Kalai (an author of this paper), it confidently produced three different answers — none of them correct. When we asked for his birthday, it gave three different dates, likewise all wrong.”

That’s not all. AI models also struggle during unprecedented market events because they learn from historical patterns rather than genuinely understanding evolving economic conditions.

Furthermore, if most investors start using the same AI models, that would only lead to crowded trades, amplifying volatility rather than reducing it. All these concerns have created a need for greater transparency and human oversight in AI-assisted investing.

Evidence suggests adoption of the technology remains uneven, with trust and perceived risk being commonly identified by researchers as key barriers.

Confidence in AI Advice Remains a Work in Progress

AI is among the few technologies that became mainstream extremely quickly. A great example of this is OpenAI’s ChatGPT, which reached its first 1 million users in just five days and then went on to hit 100 million users within two months. Most recently, ChatGPT crossed the 1 billion global monthly active app users milestone, roughly three years after launch, becoming the fastest app ever to do so.

While this data shows broader AI usage, evidence suggests that investors are also comfortable using AI for investment-related tasks.

A growing number of retail investors are now using the technology to research stocks, summarize financial reports, interpret unfamiliar financial jargon, screen for opportunities, and get a faster read on breaking news.

A survey by Investing.com of 938 US retail investors found that nearly two-thirds (62%) are already using AI tools to make informed investment decisions.

Among these, 23.6% use AI tools regularly and 27.4% occasionally. Only 11.5% have tried them once or twice. Notably, 21% of respondents said they have not yet used AI but are considering it, while 16.6% have no plans to use AI tools for investing at all.

“Arguably, there are only a handful of industries in which AI has proven as disruptive as quickly as it has in the financial industry,” said Thomas Monteiro, senior analyst at Investing.com. “This is even more pronounced for retail investors, where companies can now offer access to all types of financial-grade tools at a fraction of what they cost just a couple of years ago. As these models evolve and become increasingly relevant, usage is likely to continue growing—even among those already interacting with some type of AI investment assistant.”

An HSBC Ipsos survey of nearly 10,000 affluent and high-net-worth (HNWI) investors across 10 markets also found that finance and investing were the most common use cases for AI among wealthy individuals, ahead of even career or personal development.

Importantly, Gen Z and Millennials are adopting AI the fastest. These younger cohorts are using AI to spot risks, speed up research, or get a “second opinion” before talking to a human advisor.

Having said that, investors are considerably less comfortable allowing AI to manage their portfolios or execute trades autonomously. Concerns about response quality and data privacy are the largest barriers to broader adoption.

Per the Investing.com survey, among AI users, 38.9% are worried about incorrect or misleading recommendations, and 24.2% are worried about “herding,” where numerous investors use the same AI-generated signals.

A UK-focused Good Money Guide survey found a similar pattern: nearly seven in ten investors had used an AI platform, yet only 18.8% said they would trust tools like ChatGPT or Claude to give financial advice, compared with 60.2% who trusted established financial websites.

The need for professional advice remains across all age groups, with a hybrid approach as the ideal future decision-making model.

“Clients are increasingly using AI to explore their options, but when it comes to making investment decisions, they value judgement, context, and accountability from a trusted wealth adviser.”

– Barry O’Byrne, CEO of International Wealth & Premier Banking at HSBC

Research Shows Trust Matters More Than Technology

A new study titled “Mind over machine: A bibliometric journey into investor perceptions of AI in stock markets¹” provides a comprehensive overview of academic research examining how investors perceive AI in financial markets.

The literature also explores anthropomorphism, which involves designing AI to seem more human, as a factor that could potentially influence investor trust, though this remains an evolving research area for now.

Interestingly, in their previous research titled “Hey AI, should I buy this stock?” – Trust, technology, and the rise of the young investor, published in Dec. 2025, the authors looked into how young investors interact with AI-powered investment tools.

They found that millennials and Gen Z now rely increasingly on intelligent systems such as robo-advisors, stock screeners, and chatbots when making investment decisions. Among these young investors, trust is the key factor that determines whether they adopt and continue using AI tools. Moreover, they want AI systems to explain the reasoning behind stock recommendations because they also care about ethics, fairness, and accountability.

The latest study extends those findings, arguing that building trust through transparency and explainability is just as important as improving the technical performance of AI.

For this, the researchers didn’t test a new AI model; instead, they conducted a bibliometric analysis of over 700 English-language, peer-reviewed publications sourced from the Scopus database, covering the period between 1993 and 2025.

Using the PRISMA screening framework and the Bibliometrix package in R, the study maps the evolution of research on AI adoption in stock investment over the last thirty-two years.

The review found that the literature has expanded rapidly alongside advances in machine learning, deep learning, and robo-advisory services, identifying four dominant research themes: technical optimization of AI models, ethical and governance issues, behavioral psychology and investor trust, and practical applications in financial services.

When it comes to technical performance, AI has improved considerably in forecasting, risk management, and portfolio optimization. But while AI offers significant benefits in terms of efficiency, automation, and personalization, these do not automatically translate into adoption.

Investors remain skeptical, and that, according to the study, arises largely from perceived risks rather than technical shortcomings alone.

“Perceived risk remains a persistent counterforce,” said the authors. Trust is yet another major hurdle preventing wider adoption among investors.

Other concerns revolve around opacity, fairness, algorithmic bias, privacy issues, regulatory uncertainty, accountability, and the “black-box” nature of many AI models, which weaken investors’ assessment of the technology’s usefulness and reduce their willingness to rely on it for investment decisions.

In this sense, “trust and perceived risk are not peripheral concerns but central interpretive mechanisms explaining why technical capability alone is insufficient for sustained user acceptance in AI-driven stock markets,” the study stated.

It points to robo-advisor adoption, which is shaped less by efficiency than by a balance of trust and perceived risk. Here, human-like design and trust foster acceptance, while interaction issues and low awareness obstruct adoption.

“Investor adoption is shaped not solely by technical efficiency, but by the interplay between perceived risk, AI design features… and trust,” noted the study. “This synthesis aligns with recent empirical evidence showing that young investors’ engagement with AI-driven stock advisory tools is primarily governed by trust formation mechanisms rooted in transparency, personalization, and perceived fairness rather than algorithmic performance alone.”

The review also highlights that psychological factors play a critical role in adoption decisions.

Investors are more likely to accept AI when systems provide understandable explanations, demonstrate reliability over time, and allow meaningful human oversight.

Previous studies have also revealed a tension in human-AI interaction that can be addressed with social or human-like design elements, which may reduce uncertainty for some investors but increase behavioral biases for others.

Conversational robo-advisor interfaces that incorporate social cues, for instance, can increase affective trust, thereby raising recommendation acceptance among investors, even when the recommendation is not consistent with their actual risk profile.

According to the study, AI shouldn’t be analyzed as “technology alone” because even qualities that are apparently technical, like fairness and explainability, are shaped in part by the broader sociotechnical system, such as institutions, rules, and governance arrangements, in which the AI is embedded and expected to function.

Recent research into AI governance further demonstrates that investor trust in algorithmic systems is influenced by ethical safeguards, institutional credibility, and perceived accountability structures.

This emphasizes the point that “trust emerges from sociotechnical alignment rather than technical optimization in isolation.”

Overall, the study concludes that future progress depends not only on improving predictive accuracy but also on designing AI systems that are transparent, explainable, and aligned with investor expectations regarding accountability and control.

For AI investing tools, this means they are entering a decisive second phase, where the competitive advantage isn’t just generating better market signals, but making analysis trustworthy enough for investors to act on.

Interactive Brokers

In the financial industry, Interactive Brokers (IBKR ) stands out for moving beyond conventional brokerage functionality toward AI-assisted investing workflows.

The automated global electronic broker custodies and services accounts for ETFs, hedge and mutual funds, registered investment advisors, proprietary trading groups, and individual investors. The company processes trades in ETFs, stocks, options, futures, forex, bonds, precious metals, mutual funds, and futures contracts on over 170 exchanges worldwide.

Customers can also use its platform to trade certain cryptocurrencies through third-party providers.

Interactive Brokers’ trading platforms include IBKR Desktop, IBKR Mobile, IBKR Trader Workstation, IBKR Client Portal, and others, while its key product offerings include IBKR Lite, IBKR Pro, and IBKR Universal Account.

When it comes to AI usage, the brokerage doesn’t position it as an autonomous portfolio manager; rather, it has integrated AI across its products and trading platforms, enabling users to analyze portfolios, research securities, monitor risk, and generate trade instructions. However, every trade must be explicitly reviewed and approved by the client before being executed.

IBKR recently expanded its AI capabilities beyond the original Claude integration to include ChatGPT and Grok and broadened the range of products for which AI can generate instructions.

The company has also introduced “tools which use AI to streamline research and visualize relationships among trends, companies, and securities to give our clients actionable investment ideas,” noted Nancy Stuebe, Director of Investor Relations, during the company’s Q1 2026 earnings call.

She also noted improvements in the AI-powered chatbot, increasing both its “accuracy and coverage while enabling our reps to focus on more complex issues. We are also applying AI to further automate processes across areas such as onboarding, compliance, and other operations. Expanding the use of AI remains a priority across the firm, both to enhance the client experience and to improve internal efficiency,” Stuebe added.

With a market cap of $158.6 billion, IBKR is trading at $94.18, up 45.48% YTD and 57% over the past year, having hit a new all-time high (ATH) of $97.84 just last month, up over 653% since 2022. It has an EPS (TTM) of 2.34 and a P/E (TTM) of 39.98, with a dividend yield of 0.37%.

For Q1 2026, the company reported diluted EPS of $0.59 and $0.60 on an adjusted basis, up from $0.48 and $0.47, respectively, a year ago, on net revenue of $1.67 billion ($1.68 billion on an adjusted basis).

Commission revenue jumped 19% to $613 million, driven by higher customer trading volumes, which increased 25% for stocks, 20% for futures, and 16% for options. Net interest income rose 17% to $904 million, driven by higher average customer credit balances and margin loans, while other fees and services grew 10% to $86 million.

Execution, clearing, and distribution fees declined 12% to $106 million, due to lower regulatory fees. Recently, the SEC also eliminated the pattern day trader rule, but CEO Milan Galik expects it to “broaden the retail access, increase the trading frequency and engagement, and also liquidity in the markets.”

The company’s income before income taxes was $1.29 billion, and the pretax profit margin was 77% in Q1, both as reported and as adjusted.

During this period, IBKR’s customer accounts rose 31% to 4.75 million, while customer equity rose 38% to $789.4 billion. Both customer credits and margin loans also increased 35% each, to $168.8 billion and $86 billion, respectively.

Interactive Brokers Group reported $21.3 billion in total equity. It also announced an increase in its quarterly cash dividend from $0.08 per share to $0.0875 per share.

Conclusion

Driven by breakthroughs in large language models (LLMs) and accelerated computing, AI is completely changing the way investors research assets, monitor their portfolios, and execute trades.

The technology’s ability to process information rapidly, detect patterns, and enhance analytical efficiency has made it a valuable tool for institutional and retail investors alike. But trust in AI-generated investment advice is still lacking, as investors prioritize transparency, accountability, data quality, and human judgment over fully autonomous decision-making.

AI is unlikely to replace humans in investment altogether; instead, the future will involve a hybrid model in which AI serves as an analyst, and humans serve as decision-makers.

Click here to learn all about investing in artificial intelligence.

References

1. Phan, T. A., Tran, H. P., Ninh, T. T. & Nguyen, H. K. Mind over machine: A bibliometric journey into investor perceptions of AI in stock markets. Strategic Business Research, 100115 (2026). https://doi.org/10.1016/j.sbr.2026.100115