Large language models (LLMs) like ChatGPT are reshaping consumer interactions across industries. The credit and collection sector could be next.
Why it matters: As collection agencies, debt buyers, fintechs, banks, and credit unions seek to improve consumer engagement, LLMs offer a potential solution for more natural and effective communication.
Driving the news: Researchers at Purdue University have found that integrating AI-powered chatbots into autonomous vehicles (AVs) improves passenger-vehicle communication.
- The study suggests that ChatGPT made it easier for passengers to convey complex requests compared to existing vehicle systems. Instead of requiring button presses or rigid voice commands, LLMs allowed passengers to say things like “I’m in a hurry” or “It’s too cold,” and the vehicle adjusted accordingly.
- This marks one of the first experiments exploring how AVs can use LLMs to interpret user commands.
The collection connection: Imagine applying this kind of interaction to a collection portal or text messaging system. Instead of static prompts or dropdown menus, consumers could use conversational language to interact with collection agencies:
- “I need more time to pay” could trigger a response that offers a payment plan.
- “I can’t remember my balance” could automatically retrieve account details.
These interactions would feel more natural and less robotic, improving consumer experience and potentially leading to higher engagement and payment rates.
What’s next: Just as LLMs in AVs operate within certain parameters (e.g., traffic rules, weather conditions), in collections, these models would need to function within legal and compliance guidelines. This means interpreting consumer requests while respecting regulations like the Fair Debt Collection Practices Act (FDCPA) and other consumer protection laws.
Zoom in: Researchers caution that while LLMs have the potential to revolutionize interactions, they are still prone to “hallucinations,” or generating responses that aren’t factually accurate. In collections, this could lead to misinformation, which is why it’s crucial to pair AI-driven systems with rigorous compliance oversight.
