Keys to a successful GenAI use policy: Clear roles, training, vendor management

For all the hype surrounding generative artificial intelligence (GenAI), the technology has been met with a healthy skepticism in the compliance community. Compliance practitioners want to know: Is it safe? Can it be deployed ethically? Are the risks greater than the rewards? And what should an AI acceptable use policy contain?

Fundamentally, compliance officers want to know if GenAI tools can provide actionable insights and make business processes more efficient, without compromising data privacy, violating intellectual property rights, injecting bias into decision-making, or creating false or skewed results.

Lisa Wymer, deputy director, IT risk & compliance at the Massachusetts Bay Transportation Authority (MBTA), told an audience at Compliance Week’s AI & Compliance Summit at Boston University on Oct. 9 that using GenAI safely and ethically begins by crafting and implementing an AI acceptable use policy.