Recently, Artificial Intelligence (AI), and more specifically Large Language Models (LLMs), have become increasingly embedded in professional workflows. At MODUS Planning, Design, and Engagement (MODUS), we have seen this shift firsthand as our staff team, peers, clients, and industry partners engage with tools that use this technology in different ways and with different levels of comfort.

​

At MODUS, we approach AI with both curiosity and caution. We continue to test, evaluate, and question AI tools while considering their implications for our work, our clients, and the communities we serve. This statement outlines our current position on the responsible use of AI within our planning, engagement, and design practice.

​

Community planning, engagement, and design are fundamentally human-centred disciplines. Good work depends on listening carefully, understanding context, building relationships, navigating complexity, and making thoughtful decisions in the public interest. Trust is central to that process. While AI tools are advancing quickly, we do not believe they can replace the judgement, accountability, empathy, and relationship-building required to do this work well.

​

At the same time, AI cannot be ignored. Governments, organizations, and communities are already grappling with the opportunities and risks associated with these technologies. Research suggests AI can improve efficiency in specific administrative and analytical tasks, particularly when used to support human decision-making rather than replace it. Canadian studies have identified measurable benefits in areas such as document screening, information sorting, public-health monitoring, and administrative support workflows. In Canadian evidence-synthesis studies, human-in-the-loop AI systems reduced screening workloads by approximately 35% to 49% in conservative workflows, with some more automated systems reporting workload reductions of over 80%, though with greater risk of omissions and verification burden.¹

​

Research also identifies significant limitations and risks. Current evidence suggests productivity gains are often task-specific, dependent on organizational context, and closely tied to ongoing human oversight and verification.² Canadian research also highlights concerns related to bias, privacy, transparency, accountability, omission errors, and the growing verification burden associated with AI-generated outputs.³ Environmental concerns related to energy consumption and large-scale computing infrastructure also remain an active area of research and public debate.⁴ As a certified B Corporation and carbon neutral company, MODUS recognizes that environmental sustainability and corporate social responsibility must remain central considerations in how emerging technologies are evaluated and used within our practice.

​

At MODUS, we are particularly attentive to the relationship risks associated with AI-generated content. Community trust is built through demonstrated care, attention, responsiveness, and accountability. Automated outputs presented as thoughtful analysis or engagement can undermine that trust if they replace genuine listening, critical thinking, or professional judgement. In our experience, the strongest planning and engagement outcomes rarely emerge from speed alone. They emerge through dialogue, reflection, iteration, and the difficult but necessary work of understanding different perspectives and lived experiences. Sometimes, it’s going where the people are and bringing a meal to share. 

​

For that reason, we remain committed to face-to-face engagement, collaborative problem solving, and thoughtful analysis grounded in human expertise. We believe technology should support this work where appropriate, not replace the professional responsibility and human connection at its core.

​​

As AI technologies continue to evolve, MODUS will continue evaluating their appropriate role within our practice. In practical terms, this means:

• We prioritize human judgement, creativity, and critical thinking in all our work.

• We remain accountable for the accuracy, quality, and integrity of all work produced by our team.

• We review and verify information before it is shared externally.

​​

We do not use AI to:

• Produce final deliverables without meaningful human authorship and review 

• Upload confidential, private, or copyrighted information into open-source systems that may retain or train on that information

• Replace professional judgement, engagement processes, or decision-making responsibilities

• Analyze data without direct human review, interpretation, and accountability

​

As we continue to carefully monitor and adopt AI, we may use it to:

• Support the review of large public datasets or publicly available documentation

• Assist with administrative or organizational tasks such as sorting, tagging, or identifying themes in information

• Support minor plain-language editing or formatting tasks

• Assist with technical troubleshooting, coding support, or software learning

• Help locate source information or publicly available reference materials

​

MODUS will continue monitoring emerging research on the environmental, social, ethical, and professional implications of AI technologies. We will continue adapting our approach as standards, regulations, and evidence evolve in line with our company values.

​

Ultimately, when clients hire MODUS, they hire a team of thoughtful professionals who use technology thoughtfully and cautiously in support of high-quality work, strong relationships, and informed decision-making. 

This approach reflects our broader values, including a commitment to healthy, sustainable and resilient communities and care for collaboration, accountability and integrity. 

​

​

_______________________

References

​

1. Hamel, C. et al. (2020). An Evaluation of DistillerSR’s Machine Learning-Based Prioritization Tool for Title/Abstract Screening. BMC Medical Research Methodology, 20, 256 (reported median screening burden reductions of approximately 47% in evidence-synthesis workflows); Gates, A. et al. (2019). Performance and Usability of Machine Learning for Screening in Systematic Reviews. Systematic Reviews, 8, 278 (reported workload reductions of 35% to 49% in semi-automated workflows and over 80% in highly automated workflows); Fisher, A. et al. (2023). Automating Detection of Drug-Related Harms on Social Media. Journal of Medical Internet Research, 25.

2. Li, J. & Liu, H. (2026). Role of Complementary Capabilities in Artificial Intelligence Adoption and Productivity: Firm-Level Evidence from Canada. Canadian Public Policy, 52(S1).

3. Rahimi, S. A. et al. (2021). Systematic Scoping Review and Critical Appraisal of AI in Community-Based Primary Health Care. Journal of Medical Internet Research, 23; Office of the Privacy Commissioner of Canada (2023). Principles for Responsible, Trustworthy and Privacy-Protective Generative AI Technologies.

4. Canadian scholarly and policy literature continues to identify environmental and infrastructure impacts associated with large-scale AI systems, including energy demand, computing infrastructure, and data-centre resource consumption.