The influence of artificial intelligence on the lifecycle of M&A transactions

By using AI during the M&A lifecycle, there can be many benefits driving efficiency but these need to be balanced with an understanding from users that the implementation of this potentially game-changing technology still requires careful thought and management. The AI tools that have been used in transactions range from large language models, machine learning algorithms, data analytics and generative AI. Let’s take a look at these tools in more detail and the impact they can have on M&A transactions.

 

Forms of AI tools

Data Analytics

Data analytic tools are software and programs that have the ability to collect and analyse data about a business, its customers, and its competition that help uncover data-driven insights. M&A data analytics is all about asking and answering smarter questions. However, in order to access their full potential, these tools require exact information plus personnel who are capable of interpreting the datasets.

Generative AI

Generative AI is a form of AI that can generate high-quality text, images, and other content based on the data they were trained on. Generative AI was first introduced in the form of chatbots on websites that allowed you to have human-like conversations such as helpdesks on Amazon. This has now developed into large language models with applications such as ChatGPT that can answer questions and assist individuals working within the field of M&A. This form of AI can clearly create legal and regulatory risks however as there have been cases where it has incorporated copyrighted material without the creators’ permission. 

Large Language Models (“LLMs”):

There has been a rise in the development of LLMs, it is a form of AI that can mimic human intelligence, which uses statistical models to analyse vast amounts of data, learning the patterns and connections between words and phrases.

Machine Learning Algorithms

Machine learning algorithms are based on mathematical models and statistical approaches, which enable computers to study from data and come up with decisions or predictions without the need for explicit programming. The main purpose of machine learning algorithms is to establish or discover patterns that people use to make predictions by training algorithms to calculate more accurate results. However, the provision of ‘good’ data being fed in is essential for the algorithms to function successfully. 

 

How AI tools can be used in M&A deals

Due Diligence

The legal due diligence process can often be time consuming and requires a lot of human input. For large transactions, legal due diligence can involve the collation and analysis of hundreds, if not thousands of documents. There are AI tools out there that can help automate many tasks which usually require human intervention, for example, reviewing and analysing lengthy contracts, employee information and company finances.

Examples of large language models combines both supervised and unsupervised machine learning and this form of AI can be used in due diligence, irrespective of the size or the complexity of the task, allowing lawyers to uncover important information that may be pertinent to the deal. The software can also highlight any anomalies to the lawyer, such as missing pages in documents or variations in the wording of a clause. 

However, the use of large language models in a deal process can carry risks, as they can potentially create what is known as “hallucinations” and inaccurate results.  It is known that the output of an AI algorithm is dependent on its input, if the data inputted is biased then the algorithm will also be biased, leaving inaccurate results.

Contract Drafting

M&A deals rely heavily on the drafting of complex share and asset purchase agreements alongside additional documents to execute the deal and an increased use of generative AI in legal drafting is inevitable. This form of AI can clearly assist the drafting process by analysing the agreement to see where it departs from defined or agreed terms, provide drafting suggestions and flag any provisions that do not comply with data provided.

Some experts have stated that ChatGPT may be able to produce certain simple short form contracts and in practice, we’re aware of it already being used to support the research process and composition of deal information memorandums, but legitimate concerns remain around the evolution of its continued use due to its limitations and the complexities of M&A contracts.

Deal Sourcing, Target Valuation and Post-completion Integration

AI can be used in other stages of the M&A lifecycle such as deal sourcing by identifying potential targets by reviewing large pools of data across a vast range of resources such as company information, news articles and social media. It can also be helpful for buyers when determining the value of the target they are looking to acquire by delving into the data to find out exactly what is going on within the business at the earliest stage possible.

The risks here is that if you don’t ask the correct questions, it can create inaccurate and unreliable outputs. The key cause of this is manual errors made during data entry and which can have negative consequences if the results are used to influence decisions in M&A deals. 

Conclusion

It is clear that as AI continues to rapidly develop, the adoption of these technologies in M&A deals is potentially transformative and so likely to only increase. We can see that the use of AI tools can have clear benefits by significantly streamlining the process and creating data driven insights in the deal, but business stakeholders and their professional advisors using AI within M&A must strive to strike the right balance between reliance and careful on-going management of the deployment of this technology.

If you or your business has any corporate matters you would like some advice on, please contact our corporate team by email on [email protected].