Somewhere on a leafy British road the above sign sits in quiet absurdity. Drivers read it, ponder its existence, and accelerate away none the wiser. It's a sign about a sign that isn't in use.

But all is not as strange as it first seems. Typically, the sign is a placeholder which indicates that the variable message sign on a nearby gantry isn't working yet or is being tested.

AI regulation in the UK finds itself in similar territory: there are signs that things are afoot, but concrete signposting seems some way off. The Financial Times recently reported that an AI Bill isn't looking likely in the next King's Speech, which is due to take place in mid-May. 

Put simply, the UK is still in testing mode on AI regulation, trying to work out where it wants to go and how it wants to get there.

This is why the Digital Regulation Cooperation Forum (DRCF) has assumed such importance in shaping Britain's current approach to AI. On 10 March 2026, the forum convened its second Responsible AI Forum in which all those involved in AI could better understand the who, what, when, where, why and how of UK AI rules in 2026.

What follows is our second article on the practical lessons for senior leadership teams and general counsel who need to understand (and thrive in) the shifting world of AI regulation.

Agentic AI

With the recent publication of the ICO's Tech Futures report on Agentic AI we have some insight into the regulatory thinking around this emerging technology. While the ICO work focusses on the data protection implications of deploying AI it was interesting to hear from Ofcom, UKAI and Responsible Intelligence about trends and developments of agentic AI.

Accountability in the agentic AI supply chain is one senior leaders will be pondering before embarking on any projects. It is clear there needs to be discussion around where boundaries lie within the agentic ecosystem and the web of relationships this emerging tech inevitably brings. If something were to go wrong – and there are already many examples to choose from – how will liability be assessed and determined and how will this reflect on your brand's reputation as well as the bottom line?

Place agentic AI in the consumer context and then you significantly up the ante. Transparency and explainability will be key to compliance but also so consumers can understand exactly what it is they are signing up to and how they will be protected in line with existing laws and guidance. We were directed to several publications on agentic AI recently released by the CMA which provide insight into the regulator's thinking:

• Agentic AI and consumers - GOV.UK
• Complying with consumer law when using AI agents - GOV.UK
• AI and collusion: frontiers, opportunities and challenges - CMA

While agentic AI presents opportunities, senior leaders need to step back and take time to make informed decisions. It is important to experiment, but also build in pilot phases and extensive testing and feedback loops before launching such a product. 

It is imperative to protect consumers from harm, whether that harm is caused directly or indirectly, to avoid potential exploitation, particularly for vulnerable consumers who may not understand the possible outcome of using agentic AI, as well as ensuring competition is across the whole AI stack so consumers are able to give clear consent to transparent decision making with their best interests at heart. 

If businesses don't get this right they will likely find themselves falling foul of existing laws and potential accusations of manipulation, price discrimination and lack of competition in the goods and services offered.

Consumer trust is paramount and at the moment trustworthiness in AI, let alone agentic AI, is a huge issue. Many people are aware of fraud and misinformation but when it comes to agentic AI, unless there is transparency and explainability, as well as the right to challenge decisions, it seems the uptake will not match the hype. That said, responsible development with implemented safeguards, allowing for genuine, freely given, informed consent might see a very different landscape in 12 months' time. 

Children's wellbeing 

With the backdrop of the government's consultation on UK children's digital wellbeing, covering social media age bans, curfews, AI chatbots and gaming (for more see our article here), the DRCF hosted a session on growing up with chatbots. 

Depending on your point of view there were some fascinating/terrifying statistics shared by the expert panel, e.g. one third of UK teenagers use a chatbot for an emotional relationship, 56% of UK teenagers believe AI can think, 23% believe AI can feel emotions and 40% have no concerns about taking advice from an AI chatbot.

Children exhibit a high level of trust in chatbots, often blurring the boundary between what is a chatbot and what is a friend. While LLMs are improving all the time, they don't have empathy – they are merely learning how to respond to emotional questions, and we know they don't always get the response right.

This is why the ICO's Age Appropriate Design Code is so important and for those exploring AI tools that are in scope, it is essential to invest time and resources in getting this right. Media and regulatory scrutiny in this area are at an all time high globally and no-one wants to be making headlines for the wrong reasons. 

Well-designed chatbots, that don't process children's special category data, have safety by design features enabled, and are segmented to deal with cognitive development do have a positive role to play in children's lives.

The discussion concluded that children and parents need to upskill to understand the tech the children are being exposed to and have informed conversations around its use. Companies operating in this space need to carefully consider their legal, regulatory and ethical obligations and ensure they always put the best interests of the child first. This raises nuanced questions about revenue models, creating dependency, doom scrolls etc. but that, as they say, is for another day! 

UK Government AI roadmap

To round off the day, Mary Jones, Director of AI Strategy and Preparedness at DSIT provided an update on the UK government's AI Strategy. It may come as a surprise that 75% of the AI Opportunities Action Plan is complete one year on, with 38 out of the 50 recommendations met. 

The UK government is clear this is the foundation on which it can now build by "looking up and looking out" to get AI working across the economy and ensure responsible but faster adoption. It is hoped that the five established AI Growth Zones will be key to unlocking private investment, driving job creation and building the required data centre capacity. An additional £500 million worth of funding to back UK AI companies is also in place. 

While work is underway to upskill workers it is clear the 10 million by 2030 is an ambitious target. It is hoped the appointment of sectoral industry champions will provide impetus across various industries and make real progress towards this goal. 

A new Future of Work unit has also been established to ensure responsible AI adoption, while monitoring disruption and the impact of AI on the labour market. It is also tasked with ensuring AI boosts jobs and growth while helping workers to upskill and adapt. A wide remit and one to watch with interest!

So what?

While the DRCF highlights the importance of working collaboratively with other digital regulators and industry, there is still an air of waiting, wondering and the unknown. The sign isn't absurd, it's there as a placeholder, which is very much what the current state of play seems to reflect when it comes to AI in the UK. 

Everyone is navigating this new world together and while there are competitive pressures to realise the benefits of AI, it is essential to heed the placeholder and know that you have your strategy agreed, governance in place and appropriate use cases for whichever AI tool/s you are using.

If you'd like to continue the conversation do get in touch with your usual Lewis Silkin contact.