Understanding the tech. After finding an AI-powered tool that solves a client challenge, you must fully understand the tech, Blickstein said. He said a critical question to ask the software provider is whether the algorithm’s training data comes from a database or the law firm’s data. Additionally, law firms should ask if the firm’s data will be “co-mingled” with other organizations’ data. “Also important, all the normal rules of buying technology still apply,” Blickstein said. Prospective clients should ask about the software’s security, ease of use and install, including where the software’s data resides, he noted.
Capturing your organization’s “secret sauce.” Law firms need to leverage their data to create more innovative AI tech solutions, said Relativity chief technology officer Keith Carlson. “Firms need to think about how they can use AI while maintaining their unique edge over competitors and disrupting the industry,” Carlson wrote in an email. “To do this, firms need to better retain the key data elements of how they do business today. Once they capture this key data, then they can work with a technology partner that can enhance and then train firm-specific models that retain how an individual firm works—their secret sauce.” For example, a firm that settles many cases out of court could train a model to predict settlement amounts, Carlson explained. Additionally, firms that handle more routine work can build AI tools that automate workflows.
Adopting artificial intelligence (AI)-powered software can be nerve-racking. The financial expenditure and time spent implementing the technology into a lawyer’s workflow can be fairly high. Plus, there needs to be human oversight over an AI platform’s results because no technology is 100% accurate and some AI tools can have biases baked into them.
But AI can be safely deployed if some prior considerations are made beforehand. From exercising due diligence in fully understanding the software to confirming how the tool will practically address a client need, AI advocates above give their top considerations before deploying AI.