When the Answer Is Free: What AI Changes About Why Members Pay Dues

Adoption is up. Value is down.

Buried in the American Society of Association Executives' first State of Associations report, released in March 2026, is a contradiction that should stop every association CEO.

ASAE reports that AI use is widespread across the sector, at 87.5 percent for content and 44.3 percent for data. In the same report, retention and engagement remain the top challenge facing associations, and the sector's financial picture is deteriorating: nearly 39 percent of CEOs report a decline against just 10 percent reporting improvement.

Read those findings together and a pattern emerges that most of the sector's AI conversation is missing. Associations are adopting AI quickly, and their core business is getting harder at the same time. Adoption is up while value is under pressure. The tools are being deployed to write newsletters faster and answer service tickets sooner, but the thing members actually pay for is eroding underneath the efficiency gains.

A note on that evidence, because we would rather you know than not. ASAE does not publish a sample size or methodology for State of Associations, and the report itself sits behind a login. The component pulse polls it draws on are small and self-selected: the November 2025 update, which puts retention and engagement at the top of the list at 31.9 percent, drew 257 responses from 4,611 invitations, a 5.1 percent response rate. Treat these as the sector's best available read on itself rather than as precise measurement.

AI competes with your core product

For most of their history, associations sold a bundle that was genuinely scarce: curated expertise, trusted answers to industry questions, access to a body of knowledge you could not easily assemble alone. That scarcity was the foundation of the value proposition and, for many associations, the quiet justification for the dues line.

Generative AI attacks that foundation directly. A member facing a regulatory question, a technical standard, or a best-practice comparison no longer has to wait for the association to interpret it. Writing in Associations Now in March 2026, Chris Vaughan argued that members are learning to solve problems without needing the association in the middle, and that associations must shift from being efficient intermediaries to being indispensable partners or become optional by default.

This is disintermediation, the same force that reshaped travel agents and encyclopedia publishers, arriving in the knowledge business. Associations are, at their core, knowledge intermediaries. When the cost of a good-enough answer drops to zero, the intermediary has to prove it still adds something the free answer does not.

Efficiency does not solve this problem. You can produce a commoditized product faster and it is still commoditized.

AI is an accelerant, not the origin

It would be a mistake, and off-key for a sector prone to alarm, to blame AI for a decline it did not start. The membership model was already under strain. Professionals belong to fewer organizations, loyalty is no longer automatic, and members increasingly weigh every dues dollar against their time and attention.

AI did not create that erosion. It removes the remaining slack and the excuse to wait. Associations that were coasting on the scarcity of information now have to answer a question they could previously defer.

That reframing matters, because it changes the executive assignment. The task is not to add AI to what you already do. It is to decide what you do that is still worth paying for, and let AI handle the rest.

A way to think about it: the Member Value Defensibility Test

The practical move is to stop treating member value as a single thing and start sorting it by how exposed each piece is. Ask one question of every benefit you offer: could a member get a good-enough version of this from a general AI tool? The answers fall into three tiers.

Commoditized by AI. Generic answers, basic how-to content, FAQ-level guidance, summaries of public information. AI does this instantly and for free, and it will keep getting better at it. Competing here is a losing position. The honest move is to stop spending scarce staff time defending it.

Contestable. Personalized guidance, curated content, active community discussion. AI can approximate the generic version, but the association can win if, and only if, its version is built on something AI cannot reach: proprietary member data, validated context, and human curation. This is the middle ground where strategy actually decides the outcome.

AI-resistant. Accredited judgment and standards, credentialing, a trusted peer community, advocacy and collective voice, proprietary benchmarking data, and human accountability for an answer. Generic AI can generate none of these. This is the durable core, and it is where investment compounds.

Used honestly, this is a reallocation tool. Most associations are still pouring effort into the commoditized tier because that is what the organization has always done.

Your moat is your data, and most organizations cannot reach it

There is one asset a general AI tool categorically lacks: it does not know your members. It cannot see who they are, what they have attended, what they have downloaded, or what they asked in your community last week. Proprietary data about your own field and your own members is the clearest source of defensible value. It is the moat.

The problem is that the moat sits largely out of reach, and the evidence on that is not encouraging.

Data Orchard's benchmark of 1,039 nonprofit organizations, drawn from 7,453 respondents across 56 countries, found that just 29 percent agree their data is complete, accurate and up to date, and only 30 percent say staff find it easy to search for and find the information they need. NTEN's 2025 survey of 220 US organizations found that silos across departments rank as the single most impactful data challenge, and that 42 percent of organizations have nobody at all with the word "data" in their job title.

Neither study is association-specific, and we are not going to pretend otherwise. But both describe the same condition, and there is no reason to think trade and professional associations are systematically better at this than the wider sector.

Vendor research points the same way, with the caveat that it comes from parties with something to sell. Momentive Software, an association management software vendor, commissioned Edge Research to survey association professionals and members in mid-2025 and reported that 39 percent of associations use AI and 40 percent have established an AI policy, with a further 20 percent drafting one. Read those numbers knowing who paid for them.

The sequence that works runs data strategy, then governance, then analytics, then AI. Most organizations are reaching for the fourth layer without having built the first three. An AI feature bolted onto fragmented, duplicated member data produces confident, wrong output, which erodes trust rather than building value. Making the proprietary-data asset usable is a precondition for competing in the contestable tier at all.

Trust becomes the product

As free content becomes abundant, its reliability becomes the scarce commodity, and that shift favors associations.

A Quinnipiac University poll of 1,397 US adults, fielded 19 to 23 March 2026 with a margin of error of plus or minus 3.3 points, found that 51 percent say they have used AI tools to research topics they are curious about, up from 37 percent in April 2025. In the same poll, 49 percent said they can trust AI-generated information only some of the time and a further 27 percent said hardly ever. Combine those two and roughly three quarters of the public is using a tool it does not fully trust.

That gap is an opening for any institution willing to stand behind an answer.

That is precisely what a credible association can do. Standards, credentialing, validated benchmarks, and the judgment of accountable experts become more valuable, not less, as the volume of unverified machine-generated content rises. The association's role shifts from being the source of the information to being the trusted, accountable layer on top of it: the body that certifies what is true, what meets the standard, and who is qualified.

Trust is not a soft benefit here. It is the product.

Governance is part of the value answer

Redefining member value forces a governance decision, because how the association itself uses AI is now part of what members are buying. If members are to trust you as the accountable layer, you have to be deliberate about your own use of AI, your handling of member data, and your disclosure of when content is machine-generated.

The strongest evidence here is not about policy documents at all. Shi and Azevedo, publishing in Nonprofit and Voluntary Sector Quarterly in 2026, surveyed 168 nonprofits and found that organizations which had never had an internal conversation about these tools were dramatically less likely to support adopting them, with an odds ratio of 0.04. The conversation, not the document, was the variable that mattered.

Governing AI is not a brake on adoption. It is the mechanism by which an association earns the trust its value proposition now depends on. A clear use-and-disclosure policy is not compliance paperwork. It is a statement to members about what they can rely on. Our guide to writing a nonprofit AI policy walks through the decisions it has to make.

What this asks of executives

The response is not a technology project. It is a strategy and leadership assignment, and it comes down to a handful of moves.

Audit your benefits against the three tiers, and stop investing effort in the value AI has already commoditized. Make your member data usable before buying more AI features, because the moat is worthless if you cannot reach it. Change the board conversation from "are we using AI" to "what member value are we defending, and how will we know it is working." Set a use-and-disclosure policy so the association can credibly be the trustworthy layer members cannot get from a chatbot. And measure member value, renewal, engagement and trust, rather than measuring AI usage, which tells you nothing about whether members still need you.

The sector's instinct is to ask how AI can make the association faster. The more consequential question is what the association is for once the answers are free.

Efficiency is table stakes. Defensibility is the executive's job.

Images on this page are AI-generated, created by the Center for Nonprofit AI.