AI should never be a gatekeeper: what Ofcom's new expectations mean for every UK business.
Author: Laurence Glen | Date published: Oct, 22, 2026, UK | Read est: 5 min read
Table of contents
What does Ofcom say about AI customer service?
Does Ofcom's AI customer service regulation apply to every UK business?
Why human support and AI need to work together
Five principles for responsible AI customer experience
What does this mean for businesses investing in AI?
Good AI customer service should make getting help easier
Looking to introduce AI without compromising customer experience?
AI should never be a gatekeeper: what Ofcom's new expectations mean for every UK business.
We've probably all been there. You contact a business, explain your problem to a chatbot and spend the next five minutes trying to speak to an actual person.
It's frustrating, particularly when the question should have been relatively easy to resolve.
And it's exactly the sort of problem highlighted in the latest Ofcom AI customer service assessment, published in September 2026.
The UK communications regulator has clarified its expectations around AI transparency and access to human support, making it clear that AI applications shouldn't become obstacles preventing telecoms customers from getting the assistance they need.
While these expectations specifically concern the telecoms sector, they raise important questions about responsible AI customer experience for businesses across the UK.
Because if AI makes customer service harder, what's the point?
What does Ofcom say about AI customer service?
In its September 2026 assessment, Ofcom explored how artificial intelligence is changing the way consumers interact with broadband, mobile and other communications providers.
Its research found that 53% of UK adult internet users had interacted with generative AI or an AI customer service chatbot during the previous 12 months.
But while some consumers welcomed AI's speed and convenience, others worried that automation could make it harder to access human support.
Ofcom subsequently clarified two expectations for telecoms providers.
Customers should know when they're interacting with generative AI.
Businesses should make it clear when a customer is dealing with an AI-powered application, rather than leaving them to work it out for themselves.
AI shouldn't create barriers to required support channels.
Customers must still be able to access the support channels telecoms providers are required to make available, rather than finding themselves trapped behind an AI chatbot.
It's a fairly straightforward principle. And one that makes sense well beyond telecoms.
Human behaviour is now one of the biggest security risks
No. And it's important to make that distinction.
The September assessment didn't introduce a new set of AI customer service regulations covering every UK business.
Ofcom regulates communications services, and its specific expectations relate to how existing telecoms consumer protection rules apply as companies introduce AI.
For businesses outside telecoms, those particular requirements don't automatically become legal obligations.
However, the principles behind them provide a useful benchmark for responsible AI deployment.
Whether you're running an online retailer, professional services business, hotel or contact centre, customers still expect transparency, appropriate support and a straightforward way to resolve problems.
So while the discussion around AI customer service regulation in the UK continues to evolve, businesses don't need to wait for new rules before improving their customer experience.
Why human support and AI need to work together
AI offers clear advantages for customer service teams.
It can answer routine questions instantly, handle enquiries outside opening hours and capture information before an employee becomes involved.
But there are situations where a conversation needs more than an automated response.
Imagine a customer calling their broadband provider about an incorrect bill.
An AI assistant might explain the charges, retrieve information or identify a straightforward issue.
That's useful.
But what if the customer disputes the charges, has already complained twice or needs an exception to the usual process?
Repeatedly directing them back to the same information only adds frustration.
This is where human support and AI need to work together.
AI should help identify the issue, gather relevant information and direct customers towards the appropriate outcome. If that requires a person, the transfer should be straightforward.
Importantly, the receiving employee should have the information they need to continue the conversation without asking the customer to start again.
Five principles for responsible AI customer experience
The Ofcom AI customer service assessment provides a useful opportunity for businesses to review how their own systems operate.
1. Make AI transparency a priority
Customers should understand when they're interacting with an AI assistant.
A simple introduction explaining what the assistant can help with is a good starting point.
AI transparency helps establish trust and avoids confusion about what the technology can and cannot do.
2. Make human support easy to access
If someone asks to speak to a person, there should be a practical route to appropriate support.
That doesn't necessarily mean an employee must be available immediately, 24 hours a day. But customers should understand their options, whether that's a transfer, callback or another accessible support channel.
3. Recognise when AI isn't the right solution
Complex complaints, sensitive circumstances and unusual requests may require greater flexibility or judgement.
Businesses should define the situations that trigger escalation, rather than expecting AI to handle every interaction.
4. Make every handover count
Having to repeat an explanation is one of the most frustrating parts of customer service.
AI can help by collecting the relevant information and, where appropriate, summarising it for the employee taking over.
A good handover should make the next stage of the conversation easier for everyone.
5. Measure customer outcomes, not just automation
How many enquiries did AI handle?
It's an important operational question, but it shouldn't be the only measure of success.
Businesses should also consider whether customers received useful answers, whether problems were resolved and how often they needed to contact the organisation again.
Responsible AI customer experience means looking at the outcome from the customer's perspective.
What does this mean for businesses investing in AI?
You just have to be clear about the role you want AI to play.
An AI virtual receptionist like Aila can answer calls, provide information, capture enquiries and help ensure customers aren't left without a response.
More advanced AI customer experience solutions can help businesses manage interactions across multiple channels and direct enquiries through appropriate workflows.
And Conversational Intelligence can highlight recurring problems, customer sentiment and opportunities to improve how teams deliver support.
The value comes from connecting those capabilities with your existing customer service operation.
That means deciding which enquiries can be resolved automatically, when employees should become involved and how customers move between the two.
Good AI customer service should make getting help easier
Ofcom's September assessment is a useful reminder that the success of AI cannot be measured purely by how much it automates.
Speed matters. Availability matters. Efficiency matters.
But so does being able to get help when you need it.
For telecoms providers, Ofcom has clarified how its existing regulatory expectations apply to AI customer service.
For other UK businesses, the wider lesson is just as relevant: build services around customers, make AI transparent and keep appropriate human support within reach.
Because the goal should always be a better outcome for the customer.
Looking to introduce AI without compromising customer experience?
Explore Focus AI's customer experience solutions and discover how intelligent automation can support your people.
Laurence Glen
IT Director
Laurence is the expert other IT leaders turn to when the pressure is on. He understands that today’s IT departments are expected to deliver more with less, protect the business, support users, and plan for what comes next, often all at once. His role is to simplify that complexity, turning technical challenges into clear strategies, practical solutions, and smoother day-to-day operations. With deep experience across service management, customer strategy, and business growth, he helps IT heads reduce noise, remove blockers, and create technology environments that make life easier for their teams and stronger for their business operations.
