Earlier this fall, I had the pleasure of joining Chris McNulty on the Polaris Pathways podcast alongside Andrew Borg from Synozur to discuss how AI is transforming the nonprofit sector. Chris is AIIM's Treasurer and a member of our Board of Directors, so the conversation felt like a natural extension of the strategic discussions we've been having about AI's role in our industry. We covered everything from governance gaps to the emerging "frontier firm" model, and I wanted to share some key takeaways particularly relevant to associations and information management practitioners.
One statistic from our conversation deserves particular attention: while 82% of nonprofits report using AI in some form, fewer than one in ten have formal policies governing that use. Associations are built on trust, member data, and institutional knowledge so this gap represents significant exposure.
The situation isn't unique to associations and nonprofits. A recent ASAE survey found that 88% of associations are already leveraging AI for content creation and 44% for data analysis, yet 86% feel either somewhat or not very prepared to navigate AI's impacts. We're enthusiastic adopters but cautious strategists.
Here's something that might surprise people outside our sector: associations are extraordinarily data rich. We hold decades of member interactions, professional development records, event histories, certification data, and community conversations. This positions us well for AI applications, but only if we treat that data as the strategic asset it is.
The question isn't whether AI can help associations. It absolutely can. The challenge is that data richness doesn't automatically translate to data readiness. Most associations are sitting on decades of inconsistent metadata, siloed systems that don't talk to each other, undocumented retention practices, and content scattered across platforms with no unified taxonomy. Your AMS holds one version of a member's history, your LMS holds another, your community platform holds a third, and your shared drives hold documents that no one has touched since 2014 but no one is authorized to delete.
This is where information management practitioners become essential. The skills that have defined our discipline (e.g., developing classification schemes, establishing metadata standards, creating retention schedules, mapping data flows, implementing access controls) are exactly what organizations need before they can deploy AI responsibly. Without this foundation, AI doesn't solve your information problems; it amplifies them. You end up with a chatbot that confidently serves up outdated policies, a search assistant that can't find content because it was never tagged properly, or a personalization engine that makes recommendations based on incomplete member profiles.
For association executives: if you're wondering why your AI pilots aren't delivering the results you expected, the answer probably isn't the technology. It's likely the information architecture underneath it.
Throughout the podcast, we kept returning to this framing: AI should multiply value, not replace people. For associations, the most promising early use cases accelerate or amplify what we already do well, like research, content curation, member engagement, education delivery.
We've been piloting AI-powered search that actually cites its sources, helping members find relevant content faster while maintaining transparency about where answers come from. We've used AI to accelerate research and streamline financial processes. In each case, the goal is augmentation. Our mission is advancing member outcomes, and AI helps us do more of that with the same resources. While early cases may not be transformative, these uses cases do allow us to focus on strategy and relationships.
One piece of advice I offered during the conversation: start with your existing platforms before building from scratch. Many association technology vendors now ship AI capabilities within their products. Your AMS, LMS, or community platform may already have features you haven't explored. Piloting within your current stack reduces cost, risk, and time to value.
This doesn't mean avoiding innovation. It means being strategic about where you invest. The August 2025 MIT report finding that 95% of enterprise GenAI projects fail to deliver measurable results wasn't an indictment of the technology—it was an indictment of how organizations approach adoption and integration. Start small, learn fast, scale what works.
Perhaps the most important point from our conversation: let AI handle transactions, but keep humans where community and trust are built. Associations exist because people want to belong to something larger than themselves. They want professional identity, peer connection, mentorship, and community.
AI can absolutely support these goals, matching members with similar interests, synthesizing discussion threads, personalizing content recommendations. But it shouldn't replace the human moments that create belonging. The handwritten note from a committee chair. The hallway conversation at conference. The mentor who remembers your career goals.
As AI becomes more capable, associations that maintain this balance will differentiate themselves. Anyone can deploy a chatbot. Not everyone can build genuine community.
The EU AI Act is now law, with phased obligations extending through 2026. Even associations operating entirely within the United States will feel ripple effects through vendors, data flows, and member expectations. Data privacy and access control are non-negotiable, particularly for organizations working with sensitive member information.
This is actually an opportunity for information management practitioners. The skills we've developed around governance, classification, retention, and access control are exactly what organizations need to deploy AI responsibly. We're not gatekeepers trying to slow progress. We're essential partners in making progress sustainable.
Looking ahead, I expect AI to integrate more deeply into core association systems. Your AMS won't just store data; it will generate personalized content and automate workflows. Your LMS won't just deliver courses; it will adapt learning paths based on individual progress. AI agents will evolve beyond simple search assistants toward genuine workflow automation.
Most interesting to me: AI facilitating community building. This includes intelligent matchmaking, conversation synthesis, and relationship mapping. For associations, AI can become a foundational layer that strengthens rather than replaces human connection.
For information management practitioners in associations, this presents an opportunity to leverage years of experience with unstructured and semi-structured data to help their organization responsibly and effectively leverage AI. And for association leaders without dedicated information management expertise on staff, now is the time to build that capability because AI will only amplify whatever information practices you already have.
The technology is changing fast, but the fundamentals haven't changed at all: good information management enables good outcomes. AI just raises the stakes.
The full episode is available at synozur.com/polaris.