The initial hype about the latest generation of artificial intelligence (AI) services such as ChatGPT has died down, but organizations are continuing to wrestle with appropriate uses of generative AI tools. A multitude of services are being developed that use AI in the background, such as Zoom meeting summarization tools and digital editing tools that learn a person’s voice and can create new audio recordings automatically.

There are serious concerns about privacy and security around the use of these services and some organizations have developed policies around AI use. Examples:

• Guidelines for Using ChatGPT and other Generative AI tools at Harvard
• Appropriate Use of ChatGPT and Similar AI Tools from UC Berkeley
• NPR guidance on the use of generative AI

While there have been a number of venues for learning about and discussing AI-related issues at Swarthmore, there are no formal guidelines yet beyond the existing acceptable use, privacy, and security policies. Before using AI-based services, we recommend that Swarthmore employees discuss their plans with ITS to insure that the tools have appropriate security and privacy protections.

Featured image generated by Midjourney prompt: “featured image for blog post on artificial intelligence security and privacy”