AI in big law: ‘It’s an exponential opportunity for growth,’ says Cleary

In the second instalment of IFLR’s “AI in big law” series, we speak to US law firm Cleary Gottlieb and its subsidiary company ClearyX.

Ilona Logvinova, director of practice innovation at Cleary Gottlieb in New York, and Carla Swansburg, CEO of ClearyX in Nova Scotia, Canada, share how the firms’ bold and strategic approach to AI is already making an impact across practice areas like M&A and financial regulation.

From off-the-shelf AI tools to tailored generative AI (GenAI) technology, they delve into how innovation is no longer a side project but a tool to unlock new business opportunities.

“It is really moving to the forefront of how we engage, interact, communicate and how we do business right and professionally,” says Logvinova.

A former technology lawyer, Logvinova’s long-standing interest in tech evolved into a focused passion for AI in 2018, when she recognised that technology was really moving to the forefront.

“In the legal space, we can benefit from it tremendously. And it's not just for us, but also for our clients and how we can empower the work that we do for [them] with AI, and achieve that best-in-class client service.”

At Cleary Gottlieb, AI transformation isn’t an experiment, but an innovative project in which the leadership team is fully invested in, notes Logvinova.

“When generative AI came to the forefront about two and a half years ago, everything changed – it presents an exponential opportunity for growth,” she tells IFLR.

With around five AI tools currently in use or being tested across its practice areas, the firm is blending third-party software with in-house built AI solutions tailored to specific use cases.

Earlier in March, Cleary announced the acquisition of Springbok AI, a generative AI product development company focusing on technology that supports legal services.

“We now have the capability, through our acquisition of Springbok AI, to build our own tech in-house designed for very specific legal use cases,” she continues.

“More broadly, we have a multi-pronged strategy where we are working with both vendors and off-the-shelf technologies while building bespoke ones.”

This dual approach is unlocking value across a wide array of practice areas, including banking, litigation, ESG, M&A and regulatory – and helping the firm stay ahead of evolving client needs.

But Cleary Gottlieb’s interest in AI goes even further. Just over three years ago, the firm launched ClearyX, a wholly owned subsidiary company designed to rethink how legal services are delivered through technology – including AI, large language models (LLMs) and GenAI.

ClearyX is essentially a legal service provider, with its unique billing method, which Swansburg explains, doesn’t include hourly rates.

She notes that part of ClearyX’s mandate is also working with other non-Cleary clients and law firms. This means that despite the firm being owned by Cleary Gottlieb, the services it offers are not exclusive to its parent company.

The team at ClearyX initially started with third-party tools like Kira AI, Ebrevia, Contract Express and Power BI.

“Those tools worked, but many of them weren’t built with in-house legal teams in mind,” Swansburg recalls.

“Many clients just needed simple, practical solutions – not complex, expensive platforms.”

That’s why about a year ago, ClearyX started developing its own tool called CORA.

“[It] stands for collaboration, organisation, review and automation,” explains Swansburg.

As an AI-driven foundational platform, CORA can be used to streamline financial services regulatory work and M&A support, among other things.

“The goal is to offer clients a tailored, centralised solution based on their specific needs, replacing fragmented tools with a unified, flexible workflow,” she adds.

The firm is expecting to launch the platform later this year, following a testing phase.

One of ClearyX’s most transformative innovations so far is using AI to support M&A transactional work.

“We started by using machine learning to speed up due diligence [but] we still maintain a people-led approach to ensure maximum quality,” says Swansburg.

She notes that to start with, the firm hired an experienced team of transaction managers and analysts to replace associates in conducting due diligence.

It also worked on more than a dozen deals before ClearyX was formally launched.

“The feedback was that the quality was as good, or better,” she recalls.

“Over time, we expanded beyond basic review by developing collaborative virtual deal rooms and integrating AI to organise and visualise data.”

That initial mindset has evolved into a tech-enabled model that leverages AI tools and visual dashboards to raise potential red flags within a transaction, highlight key clauses, and bring clarity to even the most complex deals.

“[This enables] clients to go through the review process more efficiently – often at more than 40% cost savings compared to traditional associate-driven models,” adds Swansburg.

In deals involving carve-outs or post-merger integration, ClearyX uses AI to track patterns across vast numbers of documents and build intuitive integration portals that help clients stay in control long after a deal closes.

With the fast-paced advancement of AI tools, client sentiment around AI remains nuanced. According to Logvinova, client collaboration is key when it comes to AI.

“While some are eager to adopt it, others remain cautious – we respect that variability. Our goal is to empower our work with AI in a way that aligns with each client's comfort level,” she says.

“Over the next 18 months, we expect broader adoption as more organisations integrate AI into their workflows – but we're not quite there yet.”

For ClearyX, it’s a slightly different story. As Swansburg points out, clients who chose to work with the firm pick it because they have an interest in using a lot of technology.

“But that’s not the whole answer,” she argues.

She notes that back when OpenAI launched ChatGPT, both ClearyX and Cleary Gottlieb were getting notes from clients saying they weren’t comfortable with them using GenAI.

“Fast forward to about a year ago, and we started getting questions about how we are using such tools to provide value for clients,” Swansburg continues.

“Even in that short period of time, I think there was some trust built.”

Transparency is critical to that shift. Swansburg stresses that it is crucial for both companies and law firms to be transparent about their own governance frameworks and risk mitigation strategies.

“Cleary Gottlieb has developed [a governance framework] that's now public on its website, for example,” she says. “Part of that is the heavy-duty information security vetting.”

She also explains that ClearyX clients often ask whether their own data is used to train AI systems for someone else’s project.

“The answer is no,” Swansburg says. The way the firm goes about training those tools is through a “curated context set of documents” – a context built on the data ClearyX has, without using the data sets themselves.

Logvinova also notes Cleary Gottlieb’s tools do not train the broader models.

“Anything that we use is a closed enterprise instance for our organisation only,” she says.

“We have a robust information security and risk process to make sure that we are properly handling any client confidential information and client data - that is a core and primary focus for us.”

Cleary Gottlieb is in the process of transitioning from pilots to formal AI rollouts. Logvinova says the firm’s focus is on “AI empowerment” across the organisation.

Part of the transition involved coaching junior and senior lawyers for a “mindset shift” – helping them to adapt, explore and discover new ways of leveraging AI tools.

When asked whether AI tools – and the resulting gains in speed and cost-efficiency – translate into more business for the firm, Logvinova says “it’s a combination of factors”.

She argues that the workflows are changing and so are the expectations. However, while certain portions of the workflow go faster, other pieces of it that weren’t necessary before, are necessary now.

“It opens up the door to do more work with the material that you're given,” she says.

“If you have a base of output that's much broader – even if that output is provided by AI – you still need to analyse it. Those human analytics are time intensive.”

Over the next year, Logvinova expects to see more law firms building their own AI tools.

“It won't just be tech that's being offered off-the-shelf, but we will see more individualised formation of the assistive technologies that are most helpful to those lawyers, being built by those lawyers,” she concludes.

To read the other articles from the "AI in big law series", click here.