Artificial intelligence isn't only creating new regulatory challenges; it is changing how people find, understand and act on regulation. Alan Blanchard explores what this means for regulators and organisations responsible for authoritative information, and why clear, structured, machine-readable content and strong information governance are becoming increasingly important.
Artificial intelligence is creating a new regulatory challenge for governments and organisations around the world. But as the debate focuses on how AI itself should be regulated, another important shift is taking place.
AI is changing how people find, interpret and act on regulation and guidance. It is also changing the information organisations create and retain through the normal course of business.
This means the relationship between AI and regulation is becoming increasingly two-way. We need effective frameworks for governing the development and use of AI, but we also need to consider whether the regulatory information that people and AI systems increasingly rely upon is fit for this new environment.
For regulators and organisations responsible for authoritative information, that raises some important questions.
1. AI regulation is becoming an information-management challenge
The regulatory landscape for AI is developing rapidly.
The EU AI Act illustrates some of the complexity organisations may increasingly have to navigate. Its governance and enforcement responsibilities are shared between the European Commission’s AI Office and national authorities. The AI Office supervises and enforces rules for general-purpose AI models, while Member State market surveillance authorities supervise and enforce rules relating to AI systems.
Responsibilities also vary according to an organisation’s role. Providers and deployers of AI systems have different obligations, while providers of general-purpose AI models are subject to specific requirements around areas including technical documentation, transparency and risk.
For organisations operating internationally, this sits alongside emerging approaches to AI governance in other jurisdictions.
The challenge is therefore not simply knowing that regulation exists. Organisations increasingly need to establish which requirements apply to them, in which jurisdiction, in which capacity and at which point in time.
That makes AI regulation an information-management challenge as much as a compliance challenge.
If regulatory requirements are fragmented across documents, websites and jurisdictions, keeping track of what applies becomes increasingly difficult. As the volume and complexity of regulation grows, organisations need regulatory information that is accurate and current, but also structured in ways that enable it to be found, compared and applied.
This becomes particularly important as organisations themselves begin using AI to help interpret regulatory requirements.
2. AI is changing how people interact with regulation
AI is no longer simply something organisations deploy behind the scenes. Increasingly, people are using AI to find information, understand complex subjects and decide what action to take.
A recent development in the Employment Tribunals provides an interesting example of the consequences.
In June 2026, the Presidents of the Employment Tribunals in England and Wales and Scotland issued new guidance on applications for interim relief. This was a direct response to a trend the tribunals had witnessed: an uplift in applications for interim relief from 20 a year to 20 a month. This was due to applicants to the tribunals using AI to research their cases and AI highlighting the possibility of making an application for interim relief.
The guidance is intended to set out very clearly what interim relief is, the high threshold that applies and the circumstances in which an application can be made so that AI systems stop recommending it as a viable option to potential applicants.
The example raises a much wider question for regulators and other organisations responsible for authoritative information: what happens when AI becomes an intermediary between regulation and the people expected to understand it?
Traditionally, regulatory guidance has largely been published on the assumption that a person will search for it, find the relevant document and read it.
Increasingly, that person's first interaction may instead be with an AI assistant.
That changes the information environment.
It means authoritative information needs to be understandable by people, but increasingly it also needs to be discoverable and interpretable by machines.
Publishing a PDF on a website may satisfy the requirement to make information publicly available, but that does not necessarily make the underlying information easy for an AI system to identify, contextualise or connect with related requirements.
Clear language remains important. But so do structure, metadata, meaningful relationships between information, consistent terminology and machine-readable formats.
This isn't about publishing information for AI instead of people. It is about recognising that people and machines are increasingly part of the same information ecosystem.
The better the quality and structure of authoritative information within that ecosystem, the greater the opportunity for AI to help people find and use it appropriately.
3. AI is changing information governance within organisations
There is another side to this information challenge.
AI doesn't just consume information. It creates it.
Meeting transcripts, summaries, prompts, actions, generated documents and AI-assisted analysis can all become part of an organisation's information environment.
That can have consequences which are easily overlooked.
Consider an AI notetaker invited to an internal meeting. Once the formal discussion has finished, participants may continue talking in less formal manner without thinking about the fact that the AI tool is still present. Their conversation could continue to be recorded, transcribed, retained or distributed.
Similarly, an employee using generative AI to help draft sensitive feedback may concentrate on the final document while paying much less attention to what they entered into the prompt.
Yet those interactions should not automatically be regarded as temporary conversations with a machine.
The Information Commissioner's Office makes the principle particularly clear for public authorities using AI: where prompts and AI-generated information are retained, they can constitute recorded information for Freedom of Information purposes, and organisations should apply appropriate information-management and retention practices.
There are also implications for personal information. Individuals have rights of access to personal information held about them, meaning organisations need to understand what information their AI tools create, where it is retained and how it is governed.
A useful principle is therefore to treat workplace interactions with AI as business communications on a par with email correspondence.
Before entering information into an AI tool, employees should consider whether they would be comfortable with that information forming part of the corporate record.
For organisations, that means AI governance cannot sit solely with the technology team. It touches information management, legal, compliance, HR, data protection and leadership.
Policies need to address not only which AI tools employees can use, but what information can be entered, what is retained, who can access it and how those records are managed throughout their lifecycle.
Regulation in an AI-enabled information environment
Much of today's AI debate understandably focuses on models, technology and risk.
But behind all three sits information.
Regulators need to publish increasingly complex requirements clearly and consistently. Organisations need to identify which requirements apply to them. AI systems need authoritative information from which to generate useful answers. And organisations need to govern the new information that AI itself creates.
This is why the future of regulation cannot be separated from the future of information management.
As AI becomes both an object of regulation and a tool through which people understand and act on regulation, the quality of the information surrounding it becomes increasingly important.
Regulatory information needs to be accurate and current. But increasingly it also needs to be clear, accessible, structured and machine-readable.
The organisations that recognise that shift will be better placed not simply to comply with AI regulation, but to operate effectively in an environment where people and machines increasingly consume the same authoritative information.
Blog post written by Alan Blanchard, TSO Business Development Director.
Frequently asked questions
How is AI changing the way people interact with regulation?
People are increasingly using AI tools to find, interpret and understand regulatory information and guidance. This means AI can become an intermediary between authoritative information and the person using it, increasing the importance of publishing regulatory content in clear, accessible and machine-readable formats.
What does machine-readable regulation mean?
Machine-readable regulation is regulatory information structured so that digital systems and AI can more easily identify, process and interpret its meaning. This can include structured content, metadata, consistent terminology and clearly defined relationships between different pieces of regulatory information.
Why is structured information important for AI?
AI relies on the quality and context of the information available to it. Well-structured information can make authoritative content easier for machines to discover, interpret and connect, while also helping people find and use the information they need.
How does AI affect information governance within organisations?
AI tools can create and retain new forms of business information, including prompts, meeting transcripts, summaries and generated content. Organisations therefore need clear governance covering what information can be entered into AI systems, what is retained, who can access it and how those records are managed.
How can regulators make information more AI-ready?
Regulators can start by ensuring authoritative information is accurate, current, accessible and consistently structured. Using metadata, taxonomies, meaningful cross-referencing and machine-readable formats can make regulatory information easier for both people and AI-enabled systems to find, understand and use.
