How artificial intelligence is transforming financial risk management

A Toronto startup shows how automated due diligence will affect not just what you invest in but who you work with

Ghostly Artificial Intelligence holding a compass

(Iker Ayestaran)

When an investment firm requested a search of documents related to Sino-Forest Corp., the Chinese forestry giant once listed on the Toronto Stock Exchange, a due diligence software engine dubbed The Brain came up with an article that asserted the company didn’t appear to own as much land as it claimed to. Unfortunately for Sino-Forest’s shareholders, this search took place years after the RCMP and Ontario Securities Commission accused the company of fraud in 2011. The article The Brain found, however, predated the investigation by several months. “Research analysts would have loved to have had that article years ago, but everyone missed it,” says Dan Adamson, one of The Brain’s developers.

Adamson, 40, is the president and CEO of DDIQ (Due Diligence Intelligence Quotient), an offshoot of Toronto artificial intelligence (AI) startup OutsideIQ focused on applying the technology to investment research. The firm launched The Brain in April 2014 after about a year of development, billing it as the first AI solution for fraud investigations, and anti–money laundering, anti-bribery and anti-corruption compliance undertakings, as well as third-party vendor screening.

DDIQ’s software engine is on the leading edge of what Adamson says is a new era of cognitive computing. “We’re expecting machines to do more and more for us. It’s no longer sufficient to just return a list of search results. There are millions of pages out there. People need to know the right information to extract from those pages. And they need it fast.”

How fast? The Brain can read a thousand articles a minute, says Dean Rootenberg, DDIQ’s chief data scientist. It recently found 800 phone numbers in the U.S. for an individual a British investigator was looking into, he adds. “How long would it have taken a person to do that, if they even could?”

The Brain isn’t about putting human investigators out of work so much as making them more effective, Adamson says. “It automatically gathers information from the open web, including from deep web sources—the sources that normal search engines cannot access without manual, human intervention—saving hours of research time.” Most government and regulatory websites are deep web sources, and The Brain is trained to access them automatically. It can also identify and rank whatever risks it flags.

As an example, Adamson tells of a Toronto high-net-worth investor who was considering investing about half a million dollars in an information technology company about to go public in June 2014. It was a deal his brokerage house encouraged him to make. Before finalizing the transaction, however, the investor wanted to be confident that his considerable investment was a wise decision. He turned to The Brain for help.

“On first glance, the firm had a nice presence and looked like a real company,” says Adamson. But as The Brain began to look deeper at what companies it had acquired and others it was connected to—as well as the people involved in them—it found information that was buried, probably intentionally. “It started to piece together a whole lot of negative information that wasn’t definitive but, to me, it sure as heck looked like [the top executive] had been involved in a series of pump and dumps,” Adamson says. The client didn’t invest in the public offering and fired his brokerage firm. He did think, however, that the fee he paid DDIQ was well worth the cost.

The firm charges clients a fixed monthly subscription fee, starting at $250. The price rises from there, depending on the volume of searches. No other company offers the same service, says Michael Beber, a former Canadian forensic accountant who is now president and CEO of Exiger, a global regulatory and financial crime, risk and compliance firm based in New York that has an equity stake in DDIQ. “In the compliance space, there are lots of competitors who have similar risk-related data, but it’s not put through any analytical engine,” says Beber. “Investors already have too much information. They need the right information, and that can only come from an analytics engine, which is fast and cheap, or a manual researcher, which is slow and costly.”

The automation of due diligence “is absolutely the wave of the future,” says Michael Volkov, CEO of Volkov Law Group LLC, based in Washington, D.C., and author of a white paper entitled “How to Automate Third Party Due Diligence Monitoring: Ten Steps to Success.” He thinks the spread and improvement of AI for due diligence will have a dramatic impact not only on the investment industry but also just as importantly on business operations. As the technology gets cheaper and more widely available, companies will use it to vet business partners and suppliers. It will become part of overall risk management for every organization, Volkov says. This will matter all the more as business relationships become globalized. Brand equity can evaporate due to a preventable mishap or an unsavoury affiliation—often instantaneously, thanks to the power of social media.

Whether DDIQ becomes a leader in this transition depends as much on its business plan as on the technology, Volkov says. “I think it’s fantastic, what The Brain does, but the next big step is what do you do with the information it gathers?”

Many of DDIQ’s clients are involved in deals that have international components. “The Brain uses a translation engine and can work in 29 languages,” says Adamson, who says the firm is working on adding to that number. “For some reason, Russian is often requested,” he laughs.

Adamson hopes that one day his company will be the hub for any and every inquiry into company operations. “My elevator pitch,” he says, “is that we intend to become the Google of risk.”