Personal AI Agents Could Increase Contact Center Traffic Rather Than Reduce It

Personal AI agents can initiate customer requests, but businesses may still need to complete them — potentially creating more work for contact centers.

Consumers are increasingly using AI agents to handle tasks such as cancelling subscriptions and booking restaurant reservations. Many requests can be completed online, through APIs or by interacting with a company’s own AI. But agents may lack the account context or permissions needed to finish a task, and some processes still require security checks or human assistance.

That could bring more calls into contact centers. Unlike people, AI agents can wait on hold without frustration, call repeatedly for updates and contact multiple businesses at once. They can also make calls outside normal business hours. If these behaviours become common, routine enquiries that might once have generated one call — or none — could lead to repeated contact.

Contact centers already manage seasonal surges, but the spread of personal agents could make higher volumes a more persistent challenge. Voice is also among the more expensive service channels, putting pressure on organizations to determine how to handle agent-generated calls efficiently.

One potential response is for businesses to use their own agents to manage requests from personal agents. A company’s agent may be able to apply business rules and workflows, access relevant customer information and complete tasks that an external agent cannot. Requests that require human involvement could then be passed to a representative with the interaction history and relevant case details.

Agentforce Contact Center

In their product announcement, Salesforce describes how its Service Cloud platform can support this approach. Its Agentforce Contact Center or other telephony providers can be used alongside AI agents, human representatives and customer data.

Salesforce says its Casey help agent operates across voice, web chat, SMS, WhatsApp and Apple Messages, while Agent Script can define what the agent is permitted to do. The company also says its Testing Center can simulate calls before an agent handles live interactions.

For cases that need escalation, Salesforce says Casey can provide a representative with an AI-generated summary, the conversation transcript and a recommended next step. Service Rep Assistant can then offer step-by-step guidance based on customer data and company knowledge.

Higher volumes

If personal agents begin making calls on customers’ behalf, contact centers may face higher volumes of repetitive enquiries, including outside business hours. This could accelerate investment in automation designed specifically to identify and handle AI-generated interactions, while making reliable access to customer context and clear escalation paths more important.

Businesses will also need to ensure their agents can work with third-party agents without bypassing security, consent or service policies. Automation may help absorb demand, but it will not necessarily remove customer-service work; in some cases, it could shift more of that work to the business.