While Simon Leroux was selling his first business in 2022, he had already started thinking about his next company—and how it could make the deal-making process easier.

After finding a buyer for his previous company, Gazelle, the Montreal-based entrepreneur started Optionality, an entire business focused on automating the mergers and acquisitions process. 

Optionality assesses buying and selling opportunities, doing the “heavy lifting” required to find outcomes and valuations before and during the M&A process. It announced it received investment from Inovia Capital in February 2024. Now it’s starting to commercialize.

“The M&A industry is ready for disruption because it’s highly manual and repetitive,” Leroux said in an interview with The Logic. “There’s a huge opportunity.”

Companies are increasingly using generative AI tools to help bolster their M&A process and due diligence. Sixteen per cent of M&A practitioners have adopted the technology, according to a Bain & Company survey at the start of this year, while 80 per cent of respondents said they’ll use it in the next three years. But there are risks to watch out for, too.

Some larger companies already have in-house AI software to assist with deal making. KPMG’s generative AI model, Kleo, assists with due diligence, said Stephen Charko, partner and national leader of KPMG Canada’s deal advisory analytics.

Kleo can analyze and write M&A documents that adhere to KPMG’s style and language after being trained on its content. “If we can help our clients move faster, it allows them to be more competitive in that deal,” Charko said.

Other companies Charko has advised use their own internal generative AI in various stages of deals, like to record meetings, parse data, craft Excel formulas, or as a “thought partner” to test investment hypotheses. It’s being used in all three stages of deal making—before, during and after, Charko said in an interview.

“Speed is an advantage when it comes to the competitive markets that we’re in,” Charko said, adding that buyers can make a pre-emptive bid sooner with generative AI’s help. “That’s what the seller wants to see—money on the table earlier, with confidence.”

For companies that are using their own generative AI processes to facilitate deal making, it requires a lot of data. CIBC Mellon—which offers asset servicing to institutional investors—has seen increased demand from clients for access to high amounts of data, CEO Mal Cullen told The Logic. 

Though CIBC Mellon doesn’t directly facilitate M&A deals, it works with clients like institutional investors that rely on its platforms to deliver data. But requests for huge swaths of data aren’t unique to companies looking to facilitate deals, Cullen said.

“Where people are investing in infrastructure, private equity, real estate—all the things that require a lot of unstructured data and different types of data—that’s evolving quickly,” Cullen said.

Company data provided by financial services firms like CIBC Mellon can help M&A advisers build financial models and assess opportunities. But third-party AI companies, like Optionality, need to draw up the same models without using sensitive financial information from private companies, Leroux said, because of data privacy.

Data inaccuracy, privacy risks and cybersecurity were the three biggest concerns among the more than 300 M&A practitioners surveyed by Bain & Company. 

Optionality creates synthetic data for private companies using machine learning, based on available data like the number of employees, location and industry. With the synthetic data, Leroux says it can model a company’s financials without using proprietary information.

There’s also concern among deal advisers over the traceability of information in using third-party AI for deals, Leroux said. Advisers need to know where every data point originated, given that they must back up every recommendation to their clients. 

The advice of some consultants, advisers or lawyers can be worth millions of dollars, and their qualitative opinion or interpretation of data isn’t going away, Leroux added. “It’s about super-charging rather than replacing.”

Still, the proliferation of AI across companies adds another layer to the due diligence process, KPMG’s Charko said. KPMG has introduced a new due diligence process in deals to assess how much a company is using and protecting data. Acquirers need to ensure they’re not buying a company that’s taking data from illegal places.

As more companies embrace generative AI for deal making, some say it could shake up entry-level roles. The private equity sector, for example, typically has entry-level employees working on analysis and building models, which could be replaced by AI. 

But Charko said completing tasks more efficiently gives associate-level employees time to build their networks, strengthen relationships, find new target companies and focus on value creation.

“Perhaps there’s a future where gen AI becomes much more pervasive and adopted,” he said. “Maybe that allows you to hire more efficiently, or work within your existing team more efficiently.”