According to a 2024 Reuters survey, its estimated that almost a quarter of legal professionals think legal departments shouldn’t use generative artificial intelligence (GenAI). This is largely because while AI systems can potentially transform businesses, they can also create very complex legal challenges. Conversely, as AI and its regulations continue to develop, having a legal expert in an AI department is no longer a luxury but a necessity, bringing several key benefits that go way beyond just compliance.

In this article, Enhesa AI Innovation Lead Ana Sofia Rolim examines the benefits of having unfettered access to expert legal insight while developing AI systems, with real-world examples from Enhesa’s own AI team.

The benefits of a legal expert in your AI team

1. Proactive risk mitigation

While AI systems and tools develop very quickly, AI regulation does not. A legal expert ensures that the team stays compliant with evolving laws, which prevents future non-compliance problems.

Not only are these laws still being shaped, but the regulatory bodies are struggling to keep up with technological advancements — meaning they must be continually revising and amending regulations to factor in the latest AI developments. A legal expert is equipped to interpret these regulations and can provide proper guidance. One common example is how to apply the General Data Protection Regulation (GDPR) in the context of machine learning.

A legal expert can also ensure that a comprehensive GenAI Policy is in force and being complied with by all employees, while constantly evolving such Policy to keep up with new regulations and tools.

2. The ethical use of AI

Another risk worth mentioning is the ethical problems that AI can bring. Legal experts are very much focused on compliance, but they also have a primary role in ensuring that AI systems are ethical. By belonging to the development team of a department, these experts can help ensure that AI solutions are designed with fairness, transparency, and non-discrimination in mind.

3. Bridging the technical/legal gap

Legal experts are crucial when bridging the gap between technical and legal terms. Developers are focused on optimizing AI systems, but legal experts ensure that legal considerations are integrated from the beginning.

4. Providing training data and evaluating legal outcomes

Lastly, legal experts also gather and provide quality and usable data to train models and evaluate the results of algorithms according to business needs. Concurrently, they’re able to provide advice to users on how to use them once they’re deployed.

Why in-house and not outside counsel?

At Enhesa, our tools are built in-house by our own developers for mostly in-house use, which makes the need for an internal legal expert even more pressing. There are several reasons for this, including cost efficiency and real-time collaboration.

Understanding company requirements and goals

In-house legal experts have a global understanding of the company. They quickly realize the organization’s goals and processes and, therefore, provide tailored (legal) advice aligned with the company’s vision and culture. Furthermore, these experts can foster real-time collaboration with development teams as they create more tools that require constant adjustments and testing.

Reduction of development interruption

With all this expertise, legal experts will also help with rapid decision-making, either by quickly answering legal dilemmas that arise during critical moments of the development of tools (which would have to be answered by external counsels and, therefore, oblige developers to stop their activity and wait) or by giving their insight on the company’s goals. In-house experts ensure that any question is quickly addressed and resolved.

Proactive vs. reactive problem-solving

Finally, companies that don’t have in-house legal experts will be in a cycle of reactive — rather  than proactive — problem-solving. The identification of legal problems late in the process of AI system development (such as bias or improper data use) will cause costly fixes or even lead to litigation. However, in-house legal experts will proactively address these challenges during the development phase, which reduces costs and time-to-market.

AI at Enhesa

Enhesa started developing its own AI tools in 2019 by developing a named entity recognition (NER) system. From there, we employed AI systems in several parts of our processes and in several departments — for example, using Natural Language Processing (NLP) to identify legal requirements, as well as leveraging large language models (LLMs) to summarize large pieces of legislation, policies, or regulatory guidance. These tools significantly enhanced the speed, efficiency, and accuracy of legal research and compliance tasks by automating the extraction of key information from complex legal texts. By integrating AI into these processes, Enhesa’s employees and clients can stay better informed and ensure timely and precise compliance with evolving legislation, while minimizing the potential for human error.

Now, we’re extending our AI capabilities directly to our customers, providing them with GenAI search tools, visual object recognition for compliance, predictive forecasting, and improved content recommendation. Our embedded legal expert model is needed more than ever to ensure that AI tooling has the guardrails in place to continue to deliver high quality machine-honed, but expert-led content.

Since the beginning, we quickly recognized the importance of having an in-house legal expert in our AI department to help us navigate the world of data privacy, evolving regulations, and the complex legal foundation of our software and business. As the legal expert in the AI department, my work is essential in the entire AI product lifecycle. I help:

• Define the scope of software and train AI engineers on the specifics of legal implementation
• Iteratively test development releases and act as a subject matter quality assurance gate before the product reaches production
• Manage training around the software as a subject matter expert

This proactive, embedded approach has been essential to avoid delays, mitigate risks, and maintain the trust of employees, customers, and stakeholders.

Legal insights: A vital part of AI development

As AI, and particularly GenAI, continues to reshape the business landscape, the legal challenges associated with its deployment will only grow. Companies failing to have and integrate in-house legal expertise within their AI departments will stay behind in several crucial ways and, ultimately, can significantly hinder their ability to compete in the market.

It’s the firm belief of the Enhesa AI team that the need is obvious: AI innovation and legal expertise must go hand in hand.