Successfully implementing AI in legal shouldn’t be a leap of faith. It should be a calculated and measured progression.
Establishing a sturdy, scalable foundation cultivates a fertile environment for meaningful AI-driven legal work to evolve and thrive. And solidifying those initial building blocks helps secure your organization against risks down the road. You can’t build a skyscraper if your basement’s flooding through the windows.
When global energy company AES recognized the transformative potential of generative AI as ChatGPT was taking its first tentative breaths back in 2022, they saw a bright future of energy legal work unfold in front of them. And they immediately knew they needed to get on board.
Adopting AI meant opening the door to both opportunity and risk. And for an energy company operating in a volatile, highly regulated environment, the stakes couldn’t be higher.
Relativity’s Sarah Green sat down with Raquel Rodriguez, assistant general counsel at AES Corporation, and Bobby Coppola, chief strategy officer at PLUSnxt, in a webinar that explored the nuts and bolts of AES’s AI adoption initative. They shared how both building a strong foundation and choosing a strategic partner helped them scale smarter and win big with AI.
A Lifeboat in a Complex Industry
Like most industries with a lot on the line, the energy industry at large is fraught with its own particular set of challenges. Organizations face constant pressure to contain operational costs, navigate an extremely active regulatory landscape, and respond to critical matters that can spin up and escalate quickly—from environmental incidents to high-stakes investigations where reputational damage is not an option. Layer on decades of legacy data formats, like warehouses of innumerable paper records stored indefinitely, and the legal burden becomes both immense and unforgiving.
With experience at the cutting edge of the energy sector, AES’s legal team was naturally eager for technology that could transform their operations and help alleviate these mounting pressures.
When ChatGPT emerged on the market, AES leadership clocked a momentous opportunity. Looking to capitalize on the tool’s momentum, they began creating an AI adoption program for day-to-day use at all levels of the business—from individual productivity to operations to external partnerships.
The AI initiative was helmed by assistant general counsel Raquel Rodriguez and championed by AES’s global legal team. General counsel pushed the team to quickly gain fluency so they could demystify AI for their business-minded colleagues.
The Plan
Raquel’s team began with general info sessions to speak broadly about gen AI and encourage people—who typically had no experience using AI tools—to get hands on with the technology.
They developed sixty-to-ninety-minute pre-packaged trainings explaining how gen AI works, what it is and what it isn't, how it can be used for daily tasks, and what were not appropriate use cases. They also underlined potential risks.
Within legal, the team held monthly seminars to feature new tools and use cases. They ran law school scenarios to hypothesize ways in which the business could use AI, the potential risks, and how they would advise the business in these situations. And to ensure no insight was lost, they collected statistics about AI usage with monthly surveys.
The initiative then progressed to more focused workshops within different departments to drill down into the details and try to understand their colleagues’ day-to-day activities—and how AI could be integrated to make these tasks more efficient.
“I think it was very powerful for people in the business to see legal be the face of this campaign to increase adoption,” Raquel said.
Follow the Leader
But what made AES’s legal team so well-positioned to champion this AI program?
“We're already at the intersection of governance, regulation, and risk on one hand, and then trying to be a business enabler on the other hand,” Raquel explained. “We constantly have to help the business make all these decisions and business strategy—taking into account regulatory or legal risks and how to mitigate them.”
“When you have that mindset and you're facing a new problem, it's our job to adapt, and to help the business adapt. So instead of presenting the legal department as a blocker, we created additional trust. Because if legal was adopting AI for its own use, and encouraging the business to use it, then why not? There's no reason not to take the risk.”
– Raquel Rodriguez, assistant general counsel at AES Corporation
Partnering for Success
People were quickly seeing the benefits of generative AI. But AES is a global organization. Beyond legal acting as a role model, they needed to secure consistent cross-functional involvement to successfully maintain their initiative. The more ROI they could prove, the more buy-in they could secure, and the further they could expand and evolve.
Raquel and her team knew the most effective way to achieve widespread adoption was to lean on the experts for guidance. Already a Relativity user, AES partnered with Relativity service provider partner PLUSnxt to augment its e-discovery workflows using generative AI-powered Relativity aiR.
They began by running aiR for Review on a relatively large data set of approximately 400,000 documents for production. PLUSnxt worked closely with Raquel’s team and outside counsel to build, refine, and iterate prompts.
To determine ROI for this project, AES assessed three key factors: speed, accuracy, and cost reduction—which they evaluated by back-testing aiR against a more traditional continuous active learning model. The results were compelling.
Just as important, critical documents surfaced earlier—allowing AES to make strategic decisions faster, even as the matter evolved. Conducting processing and review in the same platform meant they could handle increasing volume without adding operational burden or unexpected delays. And quality control through thorough validation helped ensure the defensibility of the results.
These kind of achievements are huge for energy organization legal teams, and it was a major win for Raquel’s team: “We felt validated. We were proud of the results, and we were excited about what we could achieve moving forward.
“Speeding up insights has been critical,” Raquel reflected. “There are times when a matter is pressing, or we have limited resources, and the fact that we can expedite review and gain insights when we need them the most through the use of AI has been very important.”
Onward and Upward
Led by the AI experts on the legal team and supported by PLUSnxt, business units across AES worked together to develop a repeatable AI model and then expanded that model within teams to augment their specific use cases. They scaled smart—starting with conservative, attainable goals and then layering new goals onto their growing framework.
And it worked. With mounting evidence supporting the efficacy and utility of generative AI and Relativity aiR—and with more and more use cases on the horizon—AES’s initative was a success.
Bobby credits AES’s action-oriented approach with their runaway success: “What I think is really awesome about Raquel and AES is that they're doers. They take ownership of what they're trying to accomplish, and then they actually put it into practice.”
A Relationship Built on Trust
For AES, long‑term AI success is grounded in partnership. As Raquel explained, partners like PLUSnxt are an extension of their legal team—bringing technical expertise, sharing knowledge rather than hoarding it, and empowering AES to become more self‑sufficient over time. “And because of that, we trust them,” she explained. “As new problems or new technologies come out, we know we can explore them together and grow that relationship.”
From PLUSnxt’s perspective, that trust comes with responsibility: “We feel an enormous sense of responsibility to keep pushing the envelope, and we are very introspective about what we're doing with technology for our clients,” Bobby said. “We make sure that we measure two, three, four, five times before it's ready for consumption by our clients.”
One Tech to Unite Them All
“One of the things I love about generative AI,” Bobby said, “is it aligns all of the stakeholders. AES wants to win their cases—at a reasonable cost. Generative AI helps push better outcomes and reduce overall costs. Their outside counsel wants to win cases for them. Generative AI actually gives them information quicker and makes them better lawyers.
“And for us, the more we're able to build relationships with partners like AES and understand their data, it leads to better results for their overall e-discovery program. Generative AI helps with all of that.”
Where Do We Grow from Here
If you are curious about innovating with AI, Raquel and Bobby have a few suggestions.
“I think you should be clear about what you're trying to accomplish,” Raquel advised. “Is it speed? Is it reducing costs? That helps guide how you prioritize the AI application. Tackle one thing at a time.” Knowing your target outcome from the beginning helps you clearly ascertain and document the impact of the tool to guide how you move forward.”
Additionally, skilled utilization of AI is a matter of discernment. Sometimes that means knowing your limits. Generative AI has an immense breadth of valuable use cases, but it’s important to recognize when the tech hasn’t quite caught up to every problem. Meet the tool where it is now, but keep a keen eye on where it’s going.
Bobby also advised straying away from “Shiny New Toy Syndrome.”
“I treat generative AI like I would a musical instrument. Most songs you can play using a fit-for-purpose guitar or piano. You don't need a different guitar or piano for every different song. It's the same thing with gen AI. It really comes down to who's playing the instrument, and how much you practice.”
– Bobby Coppola, chief strategy officer at PLUSnxt
“If you have fit-for-purpose generative AI and a foundational tool like RelativityOne—you build your foundation first, learn the tools, practice the tools, and become experts on the tools. Then you can start really tinkering around the edges. You can get a lot of bang for your buck with what you have,” Bobby said.
For the hesitant, Raquel had a few final words of advice: “I would just encourage folks to give AI a try. The legal profession isn’t disappearing anytime soon. If we evolve using these tools, we can provide more strategic, more impactful legal advice. We can leave the minutia to the AI tool and focus on the things that have the biggest impact for our clients and for the business. AI is just an opportunity to become even more effective.”
Graphics for this article were created by Sarah Vachlon.
