State Data and Information Officers Talk AI at Tech Education Conference

Artificial Intelligence continues to be a hot topic among state agencies and industry partners, including this year’s Texas Association of State Systems for Computing and Communications (TASSCC) Technology Education Conference, which saw a panel of data and information officers sharing how their respective agencies have implemented AI.

Moderated by Michael Best & Friedrich LLP partner Elizabeth Rogers, the AI panel consisted of Department of Family and Protective Services Chief Data Officer Amberle Carter, Texas Department of Motor Vehicles Data Management Officer Tammi Powell and Travis County Chief Information Security Officer Brandon Rogers.

Carter, who previously spoke about AI at TASSCC’s State of the State Conference in December, detailed her agency’s formation of a center of excellence, which consists of stakeholders from various internal departments.

“We have roles and responsibilities for each different task that applies to artificial intelligence because from a data perspective, a technology perspective, a security perspective or privacy, we all bring value to the table,” Carter said. “So, as we establish what the entire process looks like, as we evaluate artificial intelligence, we meet every other week. Once that’s a well-oiled wheel, we can meet less regularly.”

Brandon Rogers discussed AI from a cybersecurity perspective, urging agencies to examine AI technologies through the lens of scope and risk to understand the best way to manage and apply them.

“One of our biggest risks is protecting intellectual property,” he said. “Say somebody is in one of those generative bots and they’re trying to find a way to reword an email. Do we now lose protection over that document where we would have protection if it wasn’t put in there.”

Powell pointed to the lack of a clear definition for AI as a major roadblock preventing extensive AI implementation, describing the technology’s evolution as “a journey.”

“It’s actually been around for a very long time by the name of natural language processing or machine learning,” said Powell. “So, we are recognizing the AI models that have already existed that we have not yet fully implemented. And then we're understanding the generative technologies as well, and the benefits that those have, but we're open to a broader understanding of it.”