Brodeur AI Guidelines
We take a human-led approach to AI. No AI-augmented or AI-generated content is ever delivered to a client or shared externally without human scrutiny.
Six principles guide how we use AI
- 01 Continuous learning
- 02 Appropriate AI tasks
- 03 Human in the loop
- 04 Data privacy and confidentiality
- 05 Transparency
- 06 Intellectual property and ethical use
How we put AI to work, responsibly
These public-facing principles cover all of our client-facing work and deliverables. They apply to every Brodeur employee and serve as context for the vendors we work with. They are not a legal contract or a full AI policy.
1. Continuous learning
As an innovative company, we continue to experiment with and selectively adopt AI tools that genuinely improve the quality and efficiency of our work. We believe in the power of AI but do not pursue automation as an end in itself. We prioritize technology that acts as a force multiplier for our human expertise.
2. Appropriate AI tasks
AI may be used in our processes, but experts are always the final authors and strategic owners of every deliverable. We restrict AI to appropriate tasks:
- Low-to-medium risk tasks. We may use AI to support content creation, primarily to review and enhance concepts developed by our team. Examples: idea generation, language polishing, and early drafts.
- High-stakes strategy. AI is never used to make strategic decisions about your brand positioning, crisis response, or reputation management. Human judgment remains the sole driver of high-value advice and sensitive communications. Examples: crisis statements, legal communications, and regulated industry content.
3. Human in the loop
No AI-augmented or AI-generated content is ever delivered to a client or shared externally without rigorous human oversight. No unreviewed, unedited AI output is ever developed or sent to a client.
- Verification. AI outputs are treated as drafts, not authoritative facts. We independently verify all statistics, quotes, and data points against trusted primary sources.
- Quality control. Our team produces every deliverable to meet your brand voice, tone, and strategic objectives.
4. Data privacy and confidentiality
Protecting your proprietary information is a baseline requirement.
- Input governance. We do not enter non-public, sensitive, or inside information into public AI models, meaning consumer tools used through public websites or apps without enterprise data protections or no-training commitments.
- Secure workflows. We follow data-protection practices designed to minimize exposure of your intellectual property and to ensure your data is never used to train public models. Where we use enterprise AI services, we do so under contracts that prohibit training on client data and include appropriate security and confidentiality protections.
5. Transparency
We believe in being open about the technology in our workflows.
- Disclosure. We call out any use of AI that materially affects a client deliverable, for example when AI imagery or video is used in a potentially public-facing asset, and we are always happy to discuss how we use AI.
- Proactive consultation. If a project involves a significant or novel use of AI beyond routine drafting, we discuss it with you in advance to set appropriate boundaries.
6. Intellectual property and ethical use
We are committed to the responsible and ethical use of technology.
- IP protection. We take all reasonable steps to avoid using AI in ways that could infringe copyrights or replicate protected content.
- Ethical use. Our team provides continuous human validation to verify accuracy, mitigate bias, and adjust as these tools and workflows change.
Questions about how we use AI?
We’re always happy to talk through how AI fits into your program.