Moving Generative AI from Experimentation to Operation

Advanced Media & Technology partner Akiba Stern is featured throughout a Harvard Business Review article exploring the evolving landscape of generative AI (gen AI) implementation—highlighting practical strategies, new and developing applications, and key considerations for successful deployment, with a focus on the importance of robust data governance frameworks.

In the article, Akiba emphasized the critical need to address data privacy and security concerns when implementing gen AI systems within a law firm. 

“That’s issue number one for law firms,” he said. “In order to be able to run a gen AI system against client data, a law firm needs to run it in a manner that keeps the provider from having access to the data ingested by the system, whether on premises or within instances in the cloud where the provider and others cannot view that information.”

To avoid potential implementation pitfalls, Akiba emphasized that companies should establish clear parameters governing how generative AI systems access and use data.

“What’s the provenance of your data?” Akiba asked. “When you use your data, what purposes are you using it for? You have to have a solid set of rules around [the data]. Specify your requirements and obligations. That is what companies want.”

When it comes to bringing gen AI into operations, Akiba proposes that, subject to the right parameters, outsourcing the operationalization process works.

“As long as you have the right parameters, you can and should contract out,” Akiba said. “Everybody does it, even banks and large pharma companies, which are among the most highly regulated enterprises. The trend continues to be to move these types of functions to providers.”

To read the full article, please visit Harvard Business Review’s website.