Your organization or IT department picked the "enterprise" AI tool employees are supposed to use, disabled half of it and told you to be careful. Most people would rather not risk using it for anything work-related that goes beyond research and basic chats. Fair enough. This session is about what you can still do to improve your day-to-day workflows, and how to do it responsibly without resorting to personal AI tools (shadow IT) behind your employer's back. It covers ten practical tips and tricks for public sector and non-profit communicators, each one starting with a restriction you actually live with (locked settings, no uploads, no history, one approved tool, agents off, no operational policy yet beyond high-level guidance) and ending with a workaround that needs no new subscription and no permission you do not already have. Every tip is platform-neutral: it works in whichever tool your organization allows. You will learn how to move beyond simple prompt frameworks into repeatable workflows instead of starting from scratch each time, how to strip the AI slop out of an approved tool's output and align it with your organization's voice and tone, how to get real work done without pasting anything sensitive, how to analyze data with AI after hashing or de-identifying it so nothing private leaves your hands, how to use AI for visuals without adding to the slop and attach content credentials to what you publish where applicable, and how to get agent-like capability even where agentic tools are still disabled. I built the list from what I see inside the government and non-profit teams I work with, and it will reflect what has changed by January 2027. You will leave with a starter pack folder of plain-text (markdown) system instructions, skills, and knowledge blocks that you can store and paste into any tool you are allowed to use. And if everything changes between now and then, as it surely will, I'll update accordingly, of course.
Mike Kujawski is President of the Centre of Excellence for Public Sector Marketing (CEPSM), the Ottawa consulting and training firm he co-founded in 2005. For over 20 years, he has worked with governments and non-profits on strategic marketing, communications, digital engagement, and behaviour change, at the intersection of marketing for good, disruptive tech, and citizen engagement. His strategies start with people. He has run hundreds of interviews, from Deputy Ministers and CEOs to working-level teams, to surface real issues and help remove roadblocks. His current AI adoption work sits in the same place: what teams can actually use inside their own rules. His projects have addressed vaccine confidence, emergency preparedness, mental health, anti-racism, gender discrimination, the environment, and food security. Clients span nearly every federal department, plus provincial, municipal, and non-profit organizations in every province and territory, including Parks Canada, the Bank of Canada, the Canadian Nuclear Safety Commission, the City of Ottawa, United Way Canada, World Spine Care, and the University of Ottawa Heart Institute. He has advised Elections Canada since 2013 and the Office of the Conflict of Interest and Ethics Commissioner since 2020, and during the pandemic he advised the Public Health Agency of Canada through its national vaccine confidence task groups, work recognized by Canada's Chief Public Health Officer. He teaches in the Institute on Governance's Digital Executive Leadership Program and serves on the executive board of the Social Marketing Association of North America. His speaking has taken him to hundreds of events on five continents, and he runs hands-on workshops on strategic social media engagement, behaviour change marketing, countering mis- and disinformation, and practical AI workflows, all of it grounded in live client work and built for public sector and non-profit realities.