Responsible AI Use at Plan Canada

AI Policy Objective: Why Responsible AI Matters

AI Policy Objective: As a non-profit organization dedicated to advancing children’s rights and equality for girls, AI has the power to expand our ability to:

• attract, retain, and delight donors
• improve the cost-efficiency of our operations
• drive greater impact in the communities where we operate.

AI Governance & Compliance Standards

Our general position on AI is that it can be a net-positive for our organization and stakeholders but must be used in a responsible and ethical manner consistent with both our strategy and our values.4 We will use AI where it supports and enhances human decision-making, expertise, and creativity, and without diminishing the ownership and responsibilities of the humans using these tools. AI tools will be selected because there is a way for them to support our mission, not just because it is a new technology.

The following is Plan Canada’s official policy on the use of AI within the organization. These policies will also be extended to any 3rd party organizations (e.g. local partners in country offices, marketing vendors, creative partners, fulfilment services, etc.) that we work with who have access to our employee, donor, or child data, to the extent possible. These requirements will be codified with 3rd parties in writing (e.g. Master Services Agreements, Data Sharing Agreements, Letter of Intents)

• Acquisition & Approval of AI Tools

  • All AI-based tools and services used by the organization, including free tools and services, will be reviewed by the AI Committee and approved by the Digital Steerco. These tools will also adhere to our existing procurement, acceptable use, privacy and data retention policies.
  • Beyond any existing policies, all AI-based tools and services will also undergo an IT security review, data governance review, and legal review.
  • Any AI-based tool or service used in the course of our work must meet the strategic or tactical goals of the organization with a clear business case and rationale. Tools/services will not be procured if their only purpose is to serve individual curiosity/experimentation and/or there is no clear business value.
  • All known risks for an AI tool will be identified prior to approval by the Steerco.
  • These principles will be followed regardless of the cost of the tools or services, including when there are no costs associated with the use or acquisition of the tool or service.
  • When new AI features are introduced for a software product already owned and in use at Plan Canada, an AI risk assessment must be performed by the AI committee before these tools may be enabled or used.
  • Personally-owned apps or tools, such as an AI app on a personally-owned phone that is not installed or used on Plan Canada devices, may be used to support work for Plan Canada, provided that no Plan-specific material (or, only publicly available information such as content from the current website) is included in the prompts and the IT acceptable use policy requirements are met.

• Data Security & Privacy Protection

  • Employee, donor, and child information, including data as well as other formats such as media files, may be used in external AI systems or non-proprietary systems, provided the vendor:
    • has strict policies that prevent the use of our information to train or expand their AI models
    • allows us to adhere to our privacy policy (internal and external), child protection requirements, and data retention schedules.
    • has signed either our Data Sharing Agreement or a Master Services Agreement containing our data sharing schedule

• Ethical AI Usage & Bias Mitigation

  • Any content/materials built by AI systems or via service providers must be reviewed and fact-checked by a staff member before being presented as accurate
  • We will use AI tools that perform as intended and not for “off-label” purposes (e.g. using approved translation tools vs. MS Co-Pilot for formal translation services).
  • We will select AI tools that consistently produce accurate outputs

• Transparency & Accountability in AI Implementation

  • We will explain when and how we use AI products when asked by our stakeholder communities and maintain an up-to-date summary of our AI policy on the website.
  • Major publications such as the annual report and Gifts of Hope catalogue will reiterate that we do not use AI to generate images.
  • All media and content generated via the use of AI must be marked as “generated by AI” for internal tracking purposes. This can be done in the Meta data for graphics/video/text and in the document itself for written text (via a watermark). The AI tool used to generate the content must also be noted.
  • Any code generated by the use of AI will be marked as “generated by AI” in the comments of the code. The AI tool used to generate the code must also be noted.

• Commitment to Trust, Fairness & Equity

  • All uses of AI tools and services will be consistent with our mission, values, and existing guidelines and policies, including the Representation, Dignity & Safeguarding in Communications Guidelines.
  • Wholly AI-generated images depicting children and program participants are not permitted. Exceptions to this must be approved by the VP, Communications.
  • Equity & Access: We will acquire and use AI in ways that enhance equity, inclusivity, and fair access to our services.
  • Fairness & Non-discrimination: We will ensure our use of AI tools does not discriminate against any of our stakeholders, including the communities we serve, those we engage with through the recruitment process, and our funders and other partners in Canada.
  • Biases: We will use AI tools that minimize bias and ensure fair outcomes for everyone, regardless of race, ancestry, place of origin, colour, ethnic origin, citizenship, creed, sex (including pregnancy and breastfeeding), sexual orientation, gender identity, gender expression, age, record of offences, marital status, family status and/or disability.
  • Trust: We will use AI in ways that allow us to maintain trust with all the communities we support and engage with.

• Regulatory Compliance & AI Risk Management

  • The AI tools we use will adhere to all regulatory requirements in the jurisdictions that we operate in.