When AI Fails & Why Compliance Still Needs Human Expertise
Key Takeaways
- AI can produce confident but incorrect answers in compliance, meaning firms risk relying on inaccurate regulatory interpretations without realising it.
- Human expertise remains essential for interpreting complex rules like MiFID RTS 22, where context, nuance and judgement are critical.
- Compliance decisions cannot be fully automated, as accountability always sits with the firm, not the technology.
- Firms should validate AI outputs against official regulatory guidance and internal policies to avoid misreporting and regulatory breaches.
- AI is a powerful support tool, but it must be combined with human oversight to manage regulatory risk effectively.
AI can support compliance teams, but it can also hallucinate regulatory guidance. This article explains why human expertise remains essential for MiFID RTS 22 interpretation, reporting decisions, and regulator-facing accountability.
AI is making waves in compliance, but as tempting as it is to rely on AI for regulatory matters, the risks are undeniable.
Compliance isn't just about finding the right answer. Its about context, nuance, and judgement. AI can assist by drafting policies, summarising regulations, and flagging anomalies, but it doesn't interpret grey areas, adapt to regulatory changes, or defend your firm in front of regulators - YOU DO!
Take a look at the screenshot I've shared below. When asked about RTS 22 transaction reporting, ChatGPT confidently gave an answer, and got it wrong. It incorrectly stated that the Execution Within Firm (Field 59) should be left blank when a firm only transmits the order and does not execute it.
However, regulatory guidance is clear: this field should be populated in every transaction report, reflecting the individual or algorithm making the execution decision. Since RTS 22 defines execution to include reception and transmission, the person or algorithm determining which firm to transmit the order to must be recorded here. If the determination is made by an external party (e.g. the client or an employee of a company within the same group), the field should be populated with NORE.
A minor mistake Perhaps. But in compliance, precision matters, and errors like this could lead to misreporting, regulatory breaches, and hefty fines.
AI is a powerful tool, but its just that, a tool. Always double check its outputs against official regulatory guidance and internal policies. While technology is growing, we still need our human expertise to keep everything on track. Our jobs Safe for now.
In the end, AI can make our work easier, but it can't replace the critical thinking, judgement, and experience we bring to compliance.
How Qomply can help
Qomply’s Regulatory Reporting Hub combines regulatory expertise with AI, automation and data analytics to deliver scalable, audit-ready reporting intelligence that reduces errors, lowers remediation costs, and minimises operational and regulatory risk.
Covering regimes including MiFIR, EMIR Refit, SFTR, CFTC, CSA, MAS, ASIC and HKMA, Qomply also offers a fully managed service and operates globally from London.
Frequently asked questions
-
It says AI can produce plausible but wrong regulatory answers. The article argues that compliance decisions still require human judgement, context, and accountability.
-
It uses Field 59, Execution Within Firm. The article says ChatGPT incorrectly stated that the field should be left blank for a firm that only transmits an order, even though the regulatory guidance says the field must still be populated.
-
It says the field should capture the individual or algorithm making the execution decision, and NORE should be used when that determination is made by an external party such as the client or a group employee. The article uses this point to show why precise interpretation matters.
-
It treats them as serious because even a small interpretive error can lead to misreporting, regulatory breaches, and fines. The article says precision matters in transaction-reporting compliance.
-
They should validate AI output against official guidance and internal policy instead of treating the tool as the final authority. The article says AI can assist with drafting and summarising, but the firm remains accountable to the regulator.
Purpose-built solutions and services for regulatory reporting
Designed to improve accuracy, reduce risk, and support confident transaction reporting.
