In September 2024, the Law Society of England and Wales published an AI strategy that set out three long-term outcomes for how the profession would navigate the technology: AI should benefit solicitors and clients in the delivery of legal services. The profession should help shape an effective regulatory landscape. And ethical use of the technology should “support the rule of law and access to justice.”
Later in June 2026, the UK government used London Tech Week to launch AI Growth Labs, testing environments for the legal sector to develop and refine AI products in line with the profession’s existing standards. Law Society chief executive Ian Jeffery said the labs had “the potential to boost innovation by allowing legal service providers to safely test AI tools against the profession’s current robust legal standards.”
Also announced in June was a pilot to develop AI legal assistants for Crown Court proceedings to improve efficiency in case research and analysis. The pilot operates in controlled environments with transparent standards for safe and ethical use.
“Artificial intelligence has the power to transform how we live, work, and govern for the better,” UK Deputy Prime Minister David Lammy said.
That is what AI leadership looks like, and that opportunity is open to every professional association.
Why Now
In addition to the UK announcements, Canada launched AI for All last month, a national strategy to increase AI adoption from 12 per cent to 60 per cent by 2034, backed by more than $2.3 billion in federal investment.
Across Europe, the EU AI Act moves into full enforcement in August 2026, with transparency obligations for organisations deploying AI in member-facing systems. For high-risk applications, such as AI used in credentialing and education, documented governance requirements apply.
Many governments are moving in a similar direction, with regulation close behind.
For associations, the more pressing question is: what are their members already doing with AI?
A European Commission survey found that just over half of Europeans now use AI, and one in four uses it at work. Institutional policy is not keeping pace.
Accenture research found that nearly a quarter of workers are already sourcing AI tools themselves, outside company systems. “People are moving faster than their organisations,” said Matt Prebble of Accenture. “Personal productivity gains are visible, but unless workflows and operations are reinvented to scale AI, they can’t be translated at an organisational level.”
Among professional associations, an industry survey indicated that 76 per cent still don’t have a formal AI policy. Only 4 per cent have an AI training budget. Most professional bodies have no official position on how, when, or whether that is appropriate.
Associations bring something to this conversation that governments and technology companies do not. They are knowledge centres for their sectors, built on decades of understanding what their members actually do, where professional risk lives, and what ethical standards mean in practice for a specific community.
The Association’s Opportunity
Associations bring something to this conversation that governments and technology companies do not. They are knowledge centres for their sectors, built on decades of understanding what their members actually do, where professional risk lives, and what ethical standards mean in practice for a specific community.
Generic AI guidance exists in abundance. Associations are positioned to provide sector-specific guidance, grounded in professional understanding of their members’ work. What does the technology mean for a solicitor whose client data is being processed by a third-party platform, or an engineer using generative design tools in a regulated environment? Associations hold that knowledge.
They are also connectors. One member’s experience working through an AI governance question becomes sector-wide learning. This can’t be replicated elsewhere.
This plays out on two levels. The first is within the association.
AI tools are already in active use across most association teams, frequently without authorisation or guidelines. An association that establishes its own AI policy defines which tools staff are authorised to use, which member data those tools may access, and which outputs require human review before being communicated or acted upon. It sets data boundaries, distinguishing between information that can be safely processed with AI tools, such as marketing content, event descriptions, and published minutes. It also identifies what information cannot be disclosed, such as financial records and anything shared in professional confidence. It names who is responsible and sets a review schedule.
The second dimension is member-facing.
Many members are navigating the same territory without institutional support. They want to know how AI is changing their professions, what standards apply, and where the ethical and legal lines fall. An association that has worked through its own position can lead and model for its sector. It moves from following its members’ questions to helping members ask better ones. It does not need to be comprehensive to be useful.
Some associations have already set out clear positions. The British Medical Association published principles for AI in healthcare, giving its members a structured framework for responsible AI use in clinical practice. The Institute of Electrical and Electronics Engineers has published the IEEE 7000 ethics standards series, which sets out how engineers should address ethical concerns in AI system design across the profession.
Professor David Strain, chair of the BMA board of science, said: “The central issue is not what AI can do; it is who will control the data that AI learns from – and therefore who ultimately controls the knowledge generated by NHS care.”
That question has a version in every sector.
An association might publish a position statement on AI, run AI literacy programmes for practitioners, advocate on standards that protect members from AI-related risk, or build guidance documents from its accumulated expertise. All of these draw on the association’s standing as an informed, trusted voice.
Where to Begin
It’s understandable that many associations have not yet formalised a policy. There is uncertainty about where to begin and concern that writing something will be outpaced by technology within months. Associations that work through this for themselves are better placed to guide members through the same questions.
Audit current use. Survey staff anonymously to understand what AI tools are already in operation. Establish the reality before drafting the policy.
Define what’s off limits. Determine which information may and may not be processed using AI tools. Member correspondence, financial records, and anything exchanged in confidence belong off the list. Be specific.
Keep humans in the loop. Identify decisions that must always involve human review, regardless of AI output. Credentialing, member discipline, and individual-facing communications are the standard starting points.
Set a disclosure standard. Determine when members will be informed that AI has contributed to something they receive. Under the EU AI Act, this is already a legal obligation in some contexts. It is good practice in all cases.
Build in regular review. AI capabilities and regulations are both changing faster than standard policy cycles. A review cadence of six to twelve months is not excessive.
Some associations already have a position on AI. The Law Society, the BMA and the IEEE have each worked through what AI means for their members. For many, that work is still ahead of them. Associations that develop a position now have a real chance to shape how AI affects their sector.