For the past few years, business outlets have published numerous stories about failed mergers of gigantic corporations due to failed due diligence studies. In fact, Hewlett-Packard’s deal with Autonomy was one of those highly reported stories of a large corporation losing billions due to an oversight in its due diligence review.[1] In 2011, Hewlett Packard acquired Autonomy for 11.1 billion, a 64% premium over the company’s 1 billion in yearly revenue. What went wrong with this deal?[2] In its review of Autonomy, Hewlett Packard was unable to detect that the CFO and CEO of Autonomy had been cooking their books, reporting profits that it had not made. For ten months before its acquisition, Autonomy reported revenues that were within the 4 percent of analyst expectations.[3] How did Hewlett Packard miss this?
The answer lies in the gigabytes of data that Autonomy handed over to Hewlett Packard. Hewlett Packard’s attorneys had to wade through thousands of documents when conducting its due diligence study, and it is often difficult to catch every potential risk in the target company.[4] Artificial intelligence can change that and save a company from purchasing companies that have cooked its books or been hacked.[5] Artificial intelligence will allow the ability to take a global look at a company and highlight inconsistencies in the documents provided to the SEC, internally, and to acquiring company.[6]

The Due Diligence Process

The primary goal of mergers and acquisitions due diligence is to identify potential risk in the transactions.[7] The acquiring company wants full disclosure from the target company (target) to reduce its potential risk from acquiring the target while understanding the business that it is taking over.[8] As seen in Omnicare, where it refused to acquire NCS with conducting due diligence, due diligence also guides the transaction.[9] Due diligence begins when both parties sign a confidentiality agreement, and the target begins to gather relevant documents for the merger.[10] Gathering the appropriate documents can prove difficult because the documents and forms may be scattered across multiple locations or formats, which could result in important information being overlooked or lost[11].
Next, the acquiring company sends the target a checklist of the types of documents, forms, and information that it would like uploaded to a virtual data room for review.[12] Virtual data rooms offer both parties a secure online location to store and review these confidential documents.[13] Before, providing the acquiring the required documents, the target sends the forms to outside counsel for review.[14] Outside counsel is responsible for managing and protecting the sensitive information of the target.[15] Once the information has been cleared by the outside counsel, it is then provided to the acquiring company for due diligence review.

Due Diligence in The Stone Age

Currently, attorneys conducting due diligence studies, have to review hundreds or thousands of documents provided by the target company to their clients. In fact, attorneys representing a typical company worth 400 million euros in an M&A agreement, have to comb through 75 to 500 contracts during their due diligence contract review.[16] However, these companies do not have just 500 contracts but rather 5,000 to 10,000 contracts on file.[17] This means that the M&A attorneys for the acquiring company only review around 5% of the total contracts that are on file.[18] Why are they only reviewing 5%? The attorneys for the acquiring company are unable to review all of the company’s documents because the traditional due diligence process is time-consuming and expensive.[19] If attorneys took a strategic approach and increased the percentage of documents that are being reviewed, it would consume 30 to 60% of the legal fees that the company was paying.[20] Due diligence review is generally done by a team of associates, which can cost $1,200 or more a document.[21]
During the due diligence process, attorneys review documents provided by the target, to identify and allocate significant risks and determine the subsequent steps in the merger. [22]Attorneys have two categories of potential risk. The first category of risk includes pre-existing business, financial or liability.[23] The second category of potential risk flows from potential anti-assignment, change of control, or confidentiality provisions the target has entered into.[24] This stage of the process is critical for the due diligence process because any oversight at this stage could result in the acquiring company purchasing a liability.[25] It is often only after the merger that acquiring companies can find missed exclusivity, nonstandard indemnification clauses, and other unexpected obligations.

What is Artificial Intelligence?

Artificial intelligence is a branch of computer science that deals with the simulation of human intelligence processes by machines.[26] This process includes acquiring information and rules for using the information, using the rules to reach approximate or definite conclusions, and self-correction.[27] The main idea of AI is the ability of computers to do mundane and repetitive tasks that humans do. AI will not replace lawyers but will rather, will become a tool that we use to improve our efficiency.[28] In house legal departments are beginning to ask outside counsel to do work in less time and are not willing to pay for work it deems as low valued added services.[29] Corporate clients are now using AI tools for other places in the business and see its added value and are looking at law firms wondering why many of these tools are not being adopted.[30]

Artificial Intelligence to Save the Day

M&A attorneys are in need of tools that will reduce the burden of the due diligence process, and help alleviate potential liabilities.[31] Attorneys representing the target are also in need of AI tools, in that they are tasked with providing all relevant documents to the acquiring company.[32] The target is held in violation of the agreement for any relevant document that is excluded from the virtual room. In deals where there are large numbers of documents, AI tools can conduct full reviews, rather than 5%, which avoids having to only analyze only a sample.[33] Attorneys representing the target company can use trained AI tools to recognize relevant documents that are standard for M&A agreements.[34] The AI tool is trained by providing it with examples of several similar provisions found in M&A agreements.[35] Some common examples that can be used in the education of the tool include confidentiality, non-competition, infringement, indemnification, controlling governance, dispute resolution and change-of-control provisions.[36] AI tools can review near-identical documents and determine the differences in both documents.[37]
The education of the AI tool will be continual, as the language used from company to company often varies depending on the user.[38] Using AI tools without the proper education runs the risk that the system may result in overlooking documents or provisions or misinterpretation of ambiguous terms.[39] The AI tool will require training from a user to help identify and define uncommon and ambiguous terms. The AI system can then be tested for its ability to recognize and define these uncommon and ambiguous terms by granting it access to numerous examples such as contracts and documents, to test its accuracy.[40]
The implementation of AI will allow for identifying, classifying, organizing, prioritizing and highlighting documents, and determining which documents needs to be provided by the target company. [41]The current process results in the attorneys providing more documents than needed, making the job of the acquiring attorneys more rigorous when done manually.[42]
When the acquiring company is given access to the target company’s documents, it has an overwhelming task of having to organize and review the content provided.[43] AI tools can be used for the automating the highlighting and cataloging the relevant provisions of each document. While the system is highlighting and prioritizing documents, it can give weight to the more problematic sections for manual review. Using AI tools will help the attorneys working for the acquiring company, with higher efficiency and speed and lower cost I when compared to the traditional method. [44]

Kira Systems

During M&A due diligence, attorneys are provided and must review hundreds or thousands of documents that have not been organized and have been stored in any number of file formats.[45] Organizing and the converting all of the documents is only the first step in a long process. [46]However, AI tools like Kira systems can help curve the time spent on this process. Kira systems is a machine learning software (AI tool) that can identify, extract, and analyze the text in contracts, emails, etc.[47] In 2015, Kira was used in over 100 billion dollars worth of M&A transactions. [48]
Though many attorneys believe that manual review is the golden method, Kira has proven otherwise in a head to head comparison.[49] When Kira and attorneys were given the same documents and provisions for contract review analysis, the manual review was shown to be just as accurate as the AI but took significantly longer.[50] Kira offers attorneys the opportunity to conduct a more comprehensive review, by analyzing more documents in less time for less money. In fact, attorneys have reported a savings of 20 to 60% of their time in reviewing contracts.[51]
Kira allows a target company to analyze its own documents to help determine its own risks and ultimately its value to improve its negotiating position.[52] By having a global understanding of its company, it can negotiate from a more knowledgeable position.[53]

Mergers are often expensive and time consuming for both sides because of the way in which the target company stores its information. Target companies are not usually deal ready. Target companies are often reactionist to potential deals, and then put in a position that requires for them to gather emails, contracts, etc. rather quickly for the deal. Kira systems can save both the target company and acquiring company time and money by allowing for document management before a potential merger is in the air[54]. Conducting manual review would take away from the core business.[55] AI systems like Kira or VDR would allow the target company the ability to upload all documentation and data to the centralized repository.[56] This would enable smooth and easier tracking with internal parties and confidential sharing with outside counsel. Kira also allows for notification of upcoming renewal and expiration dates that exist in the contract portfolio, which allows for the target and acquiring company to have a better understanding of its obligations.


During due diligence for M&A, most of the attorney fees are generated from having to spend expensive hours reviewing documents for their clients. The clients have begun to push back against their outside counsel, requiring for fixed rates and higher value added. This provides an opportunity for automating the due diligence process, which would result in cheaper and faster transactions that allow for better management of risk. AI tools offer a potential efficient solution for classifying, organizing, and prioritizing documents being provided to the acquiring company, and provides a strategy for how the acquiring company addresses its review of these documents. AI tools also provide an opportunity for target companies to be M&A ready upfront.

[1] Jack Ciesielski, How autonomy fooled Hewlett-Packard, Fortune (Dec. 2016),
[2] Id.
[3] Angela Monaghan, Hewlett Packard offloads last Autonomy assets in software deal, The Guardian (Sep. 2016),
[4] Id.
[5] Roy Strom, Wall Street wakes up to legal AI for due diligence, The Am. Law. (Dec. 2017),
[6] Id.
[7] Dewey Ray, What is merger and acquisition due diligence?, CIO (Jun 2015),
[8] Id.
[9] Daniel Davis, Omnicare v. NCS Healthcare: A Critical Appraisal, Berk. Bus. Law J. 177, 183 (Mar. 2007).
[10] Dewey Supra, note 7.
[11] Id.
[12] Cassity Ming, The due diligence process for M&A: a complete guide, Securedocs (May 2016),
[13] Id.
[14] Id.
[15] Id.
[16] Kira Systems, HOW AI IS TRANSFORMING THE DUE DILIGENCE PROCESS, Raconteur (Oct. 2018),
[17] Id..
[18] Id.
[19] Preparing your company for sale: due diligence from a seller’s perspective, Lex. (Mar. 2012),
[20] Kira Supra, note 16.
[21] Id.
[22] Supra, note 19.
[23] Id.
[24] Id.
[25] Id.
[26] Justin Evans, Why Law Students Should Embrace Artificial Intelligence, Intellectuallyjay (Sep. 2017),
[27] Id.
[28] Id.
[29] Jennifer Smith, Law firms face fresh backlash over fees, The Wall St. J. (Oct. 2012),
[30] Id.
[31] David McLaughlin, How AI helps financial institutions perform customer due diligence, Upside (Aug. 2017),
[32] Kira, Supra note 16.
[33] Id.
[34] Id.
[35] Id.
[36] Julie Sobowale, How artificial intelligence is transforming the legal profession, ABA J. (Apr. 2016),
[37] Id.
[38] Kira, Supra note 16.
[39] Id.
[40] Cassity Supra, Note 12.
[41] Id.
[42] Id.
[43] Id.
[44] Id.
[45] Debra Weiss, DLA Piper to use artificial intelligence for M&A document review, ABA J. (Jun. 2016),
[46] Id.
[47] Kira Systems, Kira Systems’ Noah Waisberg, uncovering truths & myths for AI Bootcamp session at Legalweek New York 2018, Kira System Blog (Jan. 2018),
[48] Deloitte, Deloitte forms alliance with Kira Systems to drive the adoption of artificial intelligence in the workplace (Mar. 2016),
[49] Mark Burdon, Artificial Intelligence – An Authentic Opportunity For Mergers And Acquisitions, The Deal Room (Aug. 2016),
[50] Supra, note 48.
[51] Id.
[52] Id.
[53] Id.
[55] Id.
[56] Id.