AI is Redefining Due Diligence and Abu Dhabi Is Leading the Charge

News

1. AI is Redefining Due Diligence and Abu Dhabi Is Leading the Charge

2. XPENG Secures UAE Flying Car Permit, Recalls Thousands of EVs

3. How du is Shaping the UAE’s Connected and Inclusive AI Future

4. Research Reveals AI Front-Runners and Laggards in Europe, Middle East, and Africa

5. MoneyHash and noon Payments Partner to Boost Local Payments Across the Gulf

6. K2 Think: UAE’s Tiny Yet Powerful AI Model that Challenges the Bigger-is-Better Myth

As financial crime becomes more complex and borderless, regulators and businesses across the Middle East are under growing pressure to strengthen due diligence practices. With the region’s compliance landscape evolving rapidly, spurred by tighter scrutiny, cross-border exposure, and digital financial ecosystems — AI is no longer a future consideration. It is a present-day imperative.

Against this backdrop, UK-headquartered AI and compliance firm Themis has launched a generative AI due diligence platform, called AI Investigator, which is designed to reimagine how organisations detect, assess, and respond to financial crime risk in Abu Dhabi.

Powered by domain-specific large language models (LLMs) trained on regulatory filings, financial crime typologies, and real-world conviction data, AI Investigator delivers rapid, explainable insights across complex ownership networks, adverse media, and hidden affiliations, reducing the investigation time from weeks to minutes.

The platform’s launch in Abu Dhabi is no coincidence. As the UAE pursues its National AI Strategy 2031 and positions itself as a global centre of excellence in AI, initiatives like ADGM’s regulatory sandbox and Hub71’s innovation ecosystem have created a fertile environment for applied AI. With $3.5 billion earmarked for AI investment between 2025 and 2027 and an estimated $40 billion lost annually to financial crime in the region, compliance, regulation, and AI are becoming a defining priority for both the public and private sectors.

Why Financial Crime is so Hard to Detect

For Themis CEO Dickon Johnstone, the case for AI in compliance is rooted in the fundamental nature of criminal deception. “Organized crime groups are deliberately going to lie to you,” he said. “They’ll walk into a bank with fake documents and never reveal their links to illicit operations or offshore jurisdictions.”

This is where generative AI offers transformational capabilities. Trained on thousands of typologies and real-world case data, AI Investigator is able to surface connections and behavioral patterns human investigators might miss. “We use AI to harness hidden data connections — to visualize risk and uncover the web of secrecy criminals operate in,” Johnstone explained. “99% of our clients are legitimate. We’re focused on the ones deliberately exploiting systems and people for illegal profit.”

AI on the Ground

Yet while AI may score highly in controlled tests, real-world deployment is far more complex, cautions Hakim Hacid, Chief Researcher at the Technology Innovation Institute. “You can have the best benchmark scores in reasoning or coding, but when the model is deployed in the real world, it behaves differently. AI adoption is a journey, not a single event,” said Hacid.

To close this gap, he advocates for human-in-the-loop systems and domain-specific benchmarks. “We need human feedback to train the AI effectively. I can build the benchmark — but only experts in financial crime can provide the criteria and context needed to evaluate the model,” he added.

This is precisely how AI Investigator was developed. Themis trained its models with direct input from ex-MLROs, compliance officers, and investigators, ensuring outputs are not just technically accurate but contextually sound.

The Future of Smart Compliance

Despite its automation capabilities, Themis is clear that human expertise remains central. “The AI brings together millions of data points and highlights patterns,” said Johnstone. “But the decision to proceed or reject a client is always made by a human. The machine gives you clarity and speed; the human brings judgment and accountability.”

Hacid echoed this sentiment but also warned of the growing capabilities of illicit actors. “There is a ‘blue team’ and a ‘red team.’ Criminals are using AI too, and they don’t have the constraints of governance or regulation. That’s why we have to be smarter, faster, and more ethical in our use of AI,” he said.

Eliane Andrea, Senior Advisor at the Financial Services Regulatory Authority (FSRA) of ADGM, highlighted how regulators themselves are embracing AI. “We cannot skip the future, so we have to start building ourselves in line with the evolution of AI. From a regulatory perspective, it’s essential to align with international standards.”

She added, “AI helps us integrate and analyze data, monitor systems, and detect unusual behaviors or transactions far more effectively. It helps us build risk dashboards to mitigate threats.”

Accountability by Design

In an environment of increasing scrutiny with the UAE recently removed from the FATF grey list, AI-based tools must meet new standards of transparency and auditability.

“We’ve built transparency into the platform,” said Johnstone. “Every source in AI Investigator is transparent. Every risk score is explainable. And we’ve created an internal AI ethics committee to ensure responsible use.”

This design makes the platform not only a tool for regulated firms but a regulator-aligned system capable of withstanding internal audits and supervisory inspections.

A Turning Point in Middle East Compliance

With clients spanning financial services, professional services, private equity, real estate, and government, Themis’ AI Investigator aims to help institutions move beyond reactive compliance toward proactive risk management.

By aligning AI with human expertise and regulatory foresight, the platform is well-positioned to define the next chapter of how financial crime is fought — not just in the Middle East, but globally.

“AI is here to stay,” said Hacid in closing. “The question now is how we choose to use it — and whether we can stay ahead of those who misuse it.”