AI-Powered Due Diligence: Tech Hype or the Real Deal?
Dominik Krimpmann, PhD
Business & Technology Futurist at Accenture | Helping Companies Reimagine via Disruptive Technology
Due diligence is at the heart of any M&A transaction. However, as digitization gains ground, the steadily rising tide of data poses tough challenges for even seasoned professional service providers, making it increasingly difficult to gain critical insights and deliver maximum value for clients. By incorporating artificial intelligence (AI) into the process, practitioners hope to resolve this issue. But can “AI due diligence” really deliver the goods?
The Challenge: Ever-Increasing Data Volumes
As regular readers of this blog may recall, in April 2020, I showcased how advanced analytics can help accelerate due-diligence tasks. This month, I want to present another way of dealing with the ever-growing volumes of business-critical information that due-diligence teams have to handle. To better understand the challenge, let’s take a closer look at the process.
Traditional due diligence investigations entail uploading masses of IT, financial, legal, and governance documents to a virtual data room. This is where the teams conducting the investigation access information in order to review it and report on their findings. But the complexity and volume of this work can rise exponentially as the amount of data and scope of the investigation increase.
Because M&A transactions now have to be completed faster than ever before, due diligence teams can soon find themselves bogged down by the basic processing these vast quantities of information. As a result, they may lose sight of higher-value activities – such as performing deeper analysis and gaining critical insight.
The Solution? M&A Meets AI and ML
The old saying “the devil is in the details” is certainly true when it comes to due diligence. Possible deal-breakers may lurk undetected in the depths of today’s seemingly bottomless data pools. And due diligence teams can rapidly find themselves overwhelmed by the sheer number of details presented for scrutiny.
This is where artificial intelligence comes into the picture. By deploying AI and machine-learning tools to complement and enhance traditional due-diligence approaches, leading professional-services players aim to lighten the load on the human experts, freeing them up for value-adding tasks.
State-of-the-art AI tools can rapidly process the wealth of information stored in virtual data rooms and extract invaluable insight from it. The tech works its way through documents and data points in much the same way as a human expert would – but at supercharged speeds and with greater accuracy. Small wonder, then, that some expect AI technologies to transform the due diligence process into a more proactive and data driven operation.
Real-World AI Due Diligence
That’s the theory, but what does AI due diligence look like in practice? UK-based software vendor Luminance , for example, has developed a solution to accelerate contractual reviews. Combining supervised and unsupervised machine learning, Luminance Diligence rigorously analyzes each and every document, alerting lawyers to any anomalies and helping users uncover risks in the data.
Another notable vendor in the AI-due-diligence space is Kira Systems. The company’s award Kira machine learning software automatically converts files into machine readable form and uses defined parameters to detect key information in contracts – considerably speeding up the due diligence process.
领英推荐
The Benefits Are Attractive…
Ultra-fast document processing is certainly a major advantage. But AI due diligence promises a lot more besides. Other benefits include greater efficiency. Because AI can assess literally millions of data points in a matter of seconds, human experts spend less time extracting the relevant provisions from documents, enabling them to focus on the analysis and insights delivered by the tech .
The automation and self-learning at the heart of AI due diligence are another major plus. Once you’ve automated a task, the AI algorithms can keep increasing their accuracy, since they learn with each deployment. In other words, they become even more effective and powerful every time they’re used.
Ultimately, advocates of the tech argue, AI due diligence means better transaction advice. By providing deeper diligence and better insights, these tools deliver more informed recommendations, which clients can leverage to make better decisions and achieve better outcomes.
…But There Are Downsides
While these benefits are very appealing, let’s not lose sight of the potential issues associated with AI in a critical business context like M&A due diligence. No technology is perfect: Algorithms can fail, and machine learning can fall short of the mark, leading to errors. What’s more, if documents are not sufficiently standardized, or if they’re unclear, there’s a real danger of misrepresentation.
We should also beware of relying too heavily on AI for due-diligence tasks. Without adequate human oversight, crucial information may be missed during extraction from documents, unwittingly exposing clients to risk.
And finally, we need to consider the costs of incorporating AI into due diligence. For example, will prohibitively high software license fees make AI as costly as deploying extra human consultants, if not more so?
AI or Not AI? That Is the Question
There’s no doubt that AI can radically transform the way due diligence is performed, and that’s a very welcome development. However, caution is advisable when deciding whether and how to incorporate the tech into M&A engagements. If you’re a professional services provider thinking about embarking on this journey, it’s worthwhile taking the time to carefully weigh up the various pros and cons.
PwC Advisory | Tech Strategy & Transformation | Manufacturing
3 年Great article Dominik Krimpmann, PhD Like in many other Use Cases, #AI & #ML in M&A scenario can help in assisted decision making, improving efficiency by helping to focus on more critical documents and sections of documents, validating the consistency of data. Due to the high stakes in M&A transactions, relying solely on AI & ML would definitely require caution.