UK Government Report Warns of Legal Vacuum in AI Financial Trading

These days, there’s something unsettling about entering a bank branch and knowing that algorithms are making decisions about people’s financial futures somewhere behind the polished counters and comforting smiles. They received no votes. No one can adequately describe how they function. Additionally, no one is actually keeping an eye on them, according to a recent parliamentary report.

The UK’s financial system is at risk of “serious harm” due to the government and Bank of England’s inability to regulate artificial intelligence, according to a dire warning from the Treasury committee. It’s an odd circumstance. AI is now used by more than 75% of City businesses for everything from processing insurance claims to determining your creditworthiness for a mortgage. However, none of it is governed by any particular laws.

The committee’s members aren’t holding back. Ministers and regulators, such as the Financial Conduct Authority, have come under fire for what they describe as a “wait-and-see” strategy. It’s the kind of statement that seems reasonable in a conference setting, but when you think about the stakes, it becomes more and more irresponsible. The Bank of England and the FCA maintain that the current regulations are adequate. Companies, on the other hand, are left to determine how decades-old regulations relate to technology that wasn’t around when those regulations were created.

In a direct statement, committee chair Meg Hillier said, “Based on the evidence I’ve seen, I do not feel confident that our financial system is prepared if there was a major AI-related incident and that is worrying.” It’s difficult to disagree with her. According to the report, the people who are supposed to be in charge of the system are playing catch-up with innovations they don’t fully understand, and the system is racing forward without any safeguards.

The issues are not hypothetical. According to the report, there is a real lack of transparency regarding how AI affects financial decisions. When vulnerable customers apply for loans or insurance, they may be turned down by algorithms that they are unable to question or even understand. Accountability is another area of confusion. Who is held accountable when an AI system makes a disastrous error? The source of the data? The developer of technology? The bank that used it? The problem is that nobody seems to be completely certain.

Another concern that keeps committee members up at night is financial stability. Growing use of AI raises cybersecurity risks for the industry as a whole. For necessary services, British financial companies are becoming unduly reliant on a small number of American tech behemoths, including Google, Microsoft, and Amazon. Vulnerabilities are brought about by this concentration of power. It also brings up awkward issues of control and sovereignty. What happens if Silicon Valley-controlled servers power vital financial infrastructure in the UK?

The issue of herd behavior comes next. Despite their sophistication, AI systems have a tendency to think similarly when trained on similar data. They may all come to the same conclusions at the same time during economic shocks, which could lead to liquidity crunches or mass sell-offs. According to the report, this might “risk a financial crisis.” It is not difficult to envision. The 2008 financial crisis demonstrated the interdependence of the financial system. The risk of cascading failures increases when AI is used to make split-second decisions for several businesses at once.

Additionally, fraud is growing. AI facilitates the creation of convincing deepfake media and phishing attacks. Another risk is unregulated financial advice from ChatGPT and other AI chatbots. These systems are already being asked for investment advice. The bots have a commanding voice. They frequently make mistakes. When someone follows poor algorithmic advice and loses money, there is no accountability.

The Treasury committee prefers action over further research. They are advocating for stress tests to evaluate the City’s preparedness for market shocks caused by AI. By year’s end, they want the FCA to release useful guidelines that define clear lines of accountability and explain how consumer protection laws relate to AI. It’s a fair demand. From the outside, it seems like regulators are hoping AI will somehow self-regulate or that issues will arise gradually enough to be addressed piecemeal.

In response, the FCA made the kind of statement that businesses make when they are caught off guard. Although the committee’s report implies otherwise, it appears that they have “undertaken extensive work.” The Treasury stated that it would “strike the right balance” and that “AI champions” for financial services would be appointed. It’s the kind of language that seems proactive but makes very few commitments.

The Bank of England listed actions it has already taken, such as releasing risk assessments. They will “consider the committee’s recommendations carefully.” Translation: Take immediate action without holding your breath. This is a pattern. Regulators acknowledge risks, pledge to keep an eye on developments, and then essentially carry on as before, hoping that nothing disastrous occurs in the interim.

This is especially annoying because no one disputes the potential advantages of AI. Real benefits include quicker claim processing, improved fraud detection, and more effective services. However, they are worthless if the system as a whole is susceptible to shocks that it cannot tolerate. While releasing features to the public while resolving safety concerns in real time, Tesla encountered similar skepticism regarding autonomous driving years ago. When it’s your car, that’s okay. When the entire financial system is involved, things are different.

As this develops, it seems as though the UK is taking a chance on something too significant to jeopardize. Other nations are creating laws pertaining to AI. The AI Act exists in the EU. Algorithmic accountability is a topic of serious discussion in the United States. Britain, which has historically been at the forefront of financial regulation, is lagging behind by acting as though outdated regulations intended for human decision-makers will somehow limit intelligent machines.

The government is in a desperate attempt to “turbocharge growth” through AI innovation at the time the report is released. That makes sense. There are real economic opportunities provided by the technology. However, accepting innovation is not the same as completely giving up on oversight. While claiming to be doing the former, the UK appears to be doing the latter at the moment.

The ambiguity surrounding accountability when things go wrong is arguably the most concerning feature. There are obvious channels of accountability in traditional finance if a banker makes a bad loan. These distinctions become hazy with AI. The decision was made by the algorithm, but who wrote it? Who provided the data? Who made the decision to implement it without sufficient testing? These are not speculative inquiries. When the first significant AI-related financial incident happens, they will be crucial; according to this report, it’s a matter of when, not if.

Serious consideration should be given to the committee’s warning. Everything else in the economy is supported by the financial system. Another crisis, another bailout, or ten more years of economic stagnation could result from improper AI regulation. Ironically, revolutionary new ideas are not needed to fix this. It necessitates carrying out the duties of financial regulators, which include comprehending the risks, creating precise regulations, and holding companies responsible when they violate them. It’s simple stuff. What should worry everyone is that it isn’t happening.