Ethical AI Policy

Yulu’s Artificial Intelligence (AI) Policy

At Yulu Impact Communications, we believe in the responsible and intentional use of artificial intelligence (AI) to support our mission of advancing social and environmental progress through ethical, strategic communications. AI is a tool, not a replacement, that enables us to work more efficiently so we can deepen our impact for clients. We use AI with caution, care, and an unwavering commitment to human judgment and ethical storytelling.

Equity & Access

We recognize that the development of AI technologies has often excluded communities of color, Indigenous Peoples, and other equity-seeking groups. Yulu is therefore committed to understanding and addressing bias in AI tools. We will apply an equity lens critically, transparently, and with the goal of making ethical communications more accessible and impactful.

• Algorithmic Recruitment Guardrails: To protect against historical and systemic biases in hiring datasets, Yulu strictly prohibits the use of automated AI screening, scoring, or sorting software in our recruitment processes.
• Applicant Accessibility: Conversely, we actively welcome and support job applicants using generative AI as an accessibility, translation, or leveling tool to overcome systemic neurodiverse or language-based barriers in our application process.
• Team Tooling Equity: Yulu guarantees all team members – regardless of role or seniority – equal access to subsidized, secure, and business-grade AI tooling. 

Transparency, Attribution & Client Governance

Whenever AI is used in our work – whether for ideation, drafting, or research – its involvement must be transparent to internal and external stakeholders and rightsholders. No AI-generated content will be published or presented to clients, the media, or the public without the active involvement, review, and approval of a Yulu team member. We do not tolerate plagiarism, nor do we support the uncredited use of AI-generated content. If AI plays a material role in content development, that contribution must be noted in our working file drafts.

• Split-Service Onboarding: We recognize that clients have diverse comfort levels with LLMs. During client onboarding, we establish clear contractual boundaries. For clients who request an AI-exclusion track, we guarantee that no client documents, proprietary data, or identifying details – including company and campaign names – are ever shared with AI systems. We may still utilize AI for high-level industry research and language refinement, provided all prompts are strictly anonymized and contain zero private context.  For clients who support AI leverage, we guarantee full transparency and human-led verification at every stage.

Human-Centred Communications

AI can support our process, but it cannot replace the strategy, integrity, and lived experience that our team brings to the table. All final content must reflect the voice, values, and vision of the human author or client. Our standards for accuracy, tone, and cultural context remain unchanged. No AI-generated content will be used in client deliverables, Yulu-branded materials, or media outreach without human editing and oversight.

Fair Work & The Talent Pipeline

As a certified B Corp committed to fair work and economic equity, we actively manage the systemic labor impacts of automation.

• Preserving the Professional Pipeline: We refuse to use AI to eliminate the fundamental, administrative, or base-skill tasks that historically served as the entry-level learning grounds for emerging communicators.
• Upskilling to Strategic Quality Control: We commit to mentoring and upskilling junior talent to become strategic prompt engineers, factual editors, and ethical AI auditors, ensuring that automation expands rather than limits career-growth pathways.

Privacy, Protection & Prompt Hygiene

Yulu team members must:

• Protect confidential and client information when using AI tools;
• Always use private, closed AI instances, and never enter personal, sensitive, or proprietary data into open, consumer-grade platforms;
• Verify the data-privacy standards of any AI platform before use.

To operationalize these rules, we implement the following methodologies:

• Prompt Hygiene: We train our team to practice strict prompt hygiene. This means writing concise, highly targeted queries, utilizing placeholders for proprietary data, and regularly clearing active chat histories. This limits continuous, background token compute, lowering our data vulnerability and digital carbon footprint.
• Embedded Tool Auditing: We conduct biannual audits of the silent, embedded AI features inside our everyday software stack – including Google Workspace, Zoom, Slack, and Adobe – to ensure background data harvesting is disabled and complies with our strict privacy policies.

A Living Commitment to Environmental Accountability

This policy reflects Yulu’s values today and will evolve. We are committed to learning, questioning, and improving how we use AI in ways that uphold our B Corp mission and meet the moment responsibly.

• Tracking the Compute Footprint: Generative AI is not carbon-free. Every LLM query carries an environmental tax in energy and water usage. Yulu is committed to actively tracking our team’s digital compute footprint.
• Digital Offsetting Expansion: We pledge to offset our calculated generative AI emissions annually as a core part of our carbon reduction strategy. 

As communicators, our power lies in how we tell the truth, build trust, and create meaning. No technology will ever replace that.

Implementation Resources

To help other purpose-driven organizations, B Corps, and impact leaders operationalize these commitments within their own workflows, we have developed a downloadable, ready-to-use Universal Ethical AI System Prompt Template – designed to be customized and embedded directly into your custom GPTs, Gems, or API configurations.

###