Yesterday I had the pleasure of hosting a Webinar on the topic “Agentic Customer Service isn’t about letting AI run wild”, co-hosted by @LAKEM.
We spoke about a shared reality for cutomer service teams: while agentic AI brings opportunity, unstructured autonomy introduces levels of unacceptable risk in environments where consistency, trust, and compliance really matter.
We explored what responsible agentic AI looks like in customer service, focusing on how autonomy must be deliberately designed, not improvised. By grounding agents in process, policy, and operational context, organizations can achieve predictability, auditability, and scale.
Key takeaways!
- Why unpredictability and inconsistent outcomes are the real concern (and not AI capability!)
- How process, policy, and context enable safe, repeatable agent behavior
- Why enterprise agentic AI should operate within governed workflows, not prompts
- How shifting from tasks to outcome-based resolution improves service quality
Watch the full recording: Agentic Customer Service isn’t about letting AI run wild. | Pega
We had some great questions during the webinar which were covered by Matt, but if anyone has anything else they want to share or ask then please post it below!