ChatGPT & Generative AI: Everything You Need to Know, Part 2

The EU is working on finalizing a draft AI Act, which regulates "AI Systems" used for particular "High-Risk" purposes. Like the General Data Protection Regulation (GDPR) (before it, the AI Act will reach beyond the borders of the EU, and apply to companies in the U.S. (and elsewhere) that place or put into service AI systems on the EU market, or that use an AI system's output in the EU. While the draft AI Act has not been finalized (and there are outstanding questions about how "general purpose AI Systems" will be regulated), companies should exercise caution before using generative AI tools for any purpose that may be considered "High-Risk" under the draft AI Act. These include:

• Recruitment or selection of staff, or evaluating candidates;
• Making decisions on staff promotions and terminations, for allocating tasks to, or monitoring and evaluating performance and behavior of staff;
• Management and operation of critical infrastructure; and
• Purposes connected with establishing the creditworthiness of individuals.

Additionally, if a company is considering incorporating generative AI into a chatbot that directly interacts with individuals (e.g., for customer services purposes), the AI Act will require that individuals are provided with "AI transparency," which means that they should be informed that they are interacting with an AI System (unless it is obvious from the circumstances). It may be advisable for companies to start considering how to implement this kind of "you are speaking with a bot" transparency in a customer-friendly manner.