How we use AI when assessing applications

PacifiCan uses artificial intelligence (AI), specifically Microsoft 365 Copilot, to support the intake, processing, and administrative assessment of applications submitted to its funding programs.

• AI use and purpose: A Microsoft AI platform is used to optimize program administration, verify application completeness, categorize and organize proposals, and assist program officers with preliminary data analysis which may inform recommendations. Microsoft AI tools may assist officials by summarizing applicant-provided information, identifying missing or inconsistent information, organizing applications by program-relevant categories, and supporting preliminary analysis against established program criteria. It does not make eligibility or funding decisions.
• Human review and validation: In alignment with federal standards for responsible use of artificial intelligence, program officers and designated officials review and validate AI-generated outputs, and all funding and administrative decisions are made by humans. AI-generated analysis, assessment or information may be modified, supplemented or disregarded where appropriate. Automated systems are designed to assist human judgment and administrative discretion. Program officers use AI-generated outputs as a non-binding administrative aid. Officers review the underlying application materials, validate any AI-supported analysis against program requirements, and may modify, supplement, or disregard AI-generated information before making any recommendation or decision.
• Data usage and privacy protection: AI processing is based on information provided in client submitted applications. Information is handled in accordance with the Privacy Act, the Access to Information Act, and applicable Government of Canada policies on data protection and security. Microsoft 365 Copilot has been configured to operate within the Government of Canada’s Protected B environment.
• Fairness and Accountability: We are committed to ensuring that our administrative processes are fair, transparent, secure, and we have scrutinized tools for possible bias and discrimination.
• Safeguards and quality assurance: Safeguards include human review and validation of training and output data, use of established program criteria, privacy and security controls for Protected B information, monitoring for errors or potential bias, and quality assurance review of the outputs.
• Algorithmic Impact Assessment (AIA): In accordance with the Directive on Automated Decision-Making, an Algorithmic Impact Assessment (AIA) has been completed.
• Additional information can be found on the Government of Canada AI Register (Minimum Viable Product) - Open Government Portal.

For more information on how we protect your personal information or to inquire about the review process for your application, please contact us at 1 800-338-9378.