Guest post: Generative AI – A checklist for safely getting started

By Xavier Diokno, Consilio

During the Legalweek New York 2024 conference, Consilio conducted a survey of 129 legal professionals regarding artificial intelligence. Of those surveyed, 85% indicated that they have started various AI initiatives. This ranged from developing a strategy around adoption (24%), educating the business about AI’s benefits (28%), and learning about the possibilities of the technology (33%).

Despite the willingness to adopt AI, organizations also indicated a lack of education, expertise, and training around AI. For example, 36% of organizations stated that they are unable to adopt GenAI due to a lack of expertise in this area, while 27% indicated that they lacked the right talent or resources. Likewise, 43% of the respondents indicated that that they received minimal training, whereas 17% had received no training at all. Considering these statistics, legal teams are left asking themselves:

  • What aspects / use cases of AI should we plan to implement?
  • What can a legal team do to train its staff on AI inclusive of its risks and potential?
  • Does a legal team need a staff of AI experts?
  • What can a legal team do to ensure that using AI is safe and secure?

Start with what you know & check your work

Lawyers and legal team members may be hesitant to trust AI or see it as a threat to their jobs.  Those same legal teams may already use applications that leverage AI. These include applications that your team may already be familiar with, such as those used for legal research, legal drafting, or eDiscovery. In most cases, these applications are simply being upgraded to include a new AI-based feature. Therefore, instead of having to learn a new application, a legal team simply needs to learn how to use the new AI feature (e.g., a new way of searching or a new way of drafting a document).

AI is full of new and rapidly evolving developments that are getting better and better with each passing iteration.  It is important to take time to check any output generated by AI to ensure accuracy.  Suggest or publish guidance on uses of AI that are appropriate for your team while more formal policies are developed in coordination with your broader enterprise.

Leverage free training / education through your vendor/s

There are a lot of articles and webinars about AI on the Internet. However, if your organization already licenses an application or is interested in evaluating an application, the software vendor will likely provide free training on their product. If you do engage with a vendor, provide them with a list of questions or topics beforehand. This will help ensure that (a) the person presenting is an SME in those areas and (b) the topics presented will be related to your interests. Moreover, if your team is evaluating an application, most vendors will offer a free trial of their application.

Where can I find the right talent?

As described above, in most cases, organizations will be using AI in the context of an application that addresses their specific needs. In these cases, you may want to consider candidates that either have experience working with that specific application or with database applications—applications that search and analyze large corpuses of data. These are generally individuals who are detail-oriented and enjoy poring over an application’s technical documentation. In cases when an organization is seeking guidance on how to deploy AI across a large business, departments may need an “expert” that has experience with applying AI in various contexts. This person can work with departmental stakeholders to understand the application, identify any risks or limitations, and advise on how to effectively leverage it.

How do I ensure that I use AI in a safe and secure manner?

When evaluating an AI application, the legal technology vendor should be able to answer the following questions.

  • Where is my data going?
  • Where is my data being stored?
  • How is my data being secured?
  • How will my data be used by the application, particularly the AI component or function?
  • What are the risks and limitations of the application?

AI presents some great efficiencies and opportunities. But as with any application that analyzes data, especially confidential or proprietary information, you need to know that your data is safe and secure.

Xavier Diokno is vice president of solutions and innovation at Consilio, a global provider of eDiscovery, document review, risk management and legal consulting services.