
Smith writes that there are ‘very real security and risk issues that need to be mitigated upfront before law firms can confidently take the plunge with generative AI, as well as significant groundwork that needs to be laid to effectively use it’.
He sets out the steps firms should take to prepare for AI, how they can assess what they want to use AI for, how to make sure they make the right data available, and how to train the AI to do what the firm wants.
Smith explains that ‘an experimentation phase’ can be useful, where lawyers are given safe versions of the technology to play around with that doesn’t compromise sensitive data.