OTTAWA - The Canada Revenue Agency (CRA) is exploring the use of artificial intelligence and enhanced training to improve the accuracy of information provided by its call center agents. Melanie Serjak, an assistant commissioner at the CRA, shared these plans during a meeting with Members of Parliament on a standing committee.

Serjak indicated that the agency aims to implement a more "senior and standardized" training program for its agents. Additionally, the CRA is looking to introduce automation to enhance the quality of advice given to taxpayers.

A recent report from Auditor General Karen Hogan revealed concerning statistics regarding the accuracy of responses from CRA call center staff. After conducting calls to the agency's contact centers over a four-month period, Hogan's office found that agents answered only 17 percent of individual tax questions correctly.

In her report, Hogan noted that the CRA appears to prioritize adherence to employee schedules for shifts and breaks over ensuring the "accuracy and completeness of information" provided to callers. This raises concerns about the effectiveness of the current system in addressing taxpayer inquiries.

To address these issues, the CRA has been utilizing a virtual chatbot named Charlie, which offers automated responses to frequently asked questions. Hogan's report highlighted that taxpayers are more likely to receive accurate information from the chatbot than from a live agent.

The CRA's initiatives aim to enhance the reliability of the information provided to taxpayers, addressing the significant gaps identified in the recent audit.

The agency's focus on AI and improved training reflects a commitment to better serve the public and ensure that taxpayers receive the correct information they need.