The EU Artificial Intelligence Act (EU AI Act) is the first comprehensive AI regulation in the world – and while the UK is no longer part of the EU, many UK contact centres fall within its scope.

Milestones have already passed – including bans on prohibited AI (Feb 2025) and rules for general-purpose AI (Aug 2025). However, August 2026 is the key date many organisations will be looking to – when full enforcement of the ‘high risk’ requirements comes into play.

For contact centre leaders, this is a date that should be circled in the diary, with the coming months a chance to engage with stakeholders within your business to ensure compliance will be met.

How does this apply to UK contact centres?

The EU AI Act operates with extraterritorial reach, similar to GDPR. This means it applies if your organisation:

• Employs advisors based in the EU
• Serves EU customers
• Uses AI systems whose outputs affect people in the EU
• Supplies services to EU‑based organisations

The EU AI Act uses a risk-based framework, dividing AI systems into four categories, unacceptable risk (banned), high risk (strictly regulated), limited risk (transparency required) and minimal risk (largely unrestricted). Many contact centres sit in the high-risk zone if AI is being used in employment and workforce management.

AI will typically be classed as high risk if it influences or supports decisions related to:

• Advisor performance scoring or monitoring
• Quality assurance (QA) and coaching recommendations
• Workforce scheduling or task allocation
• Decisions affecting pay, progression or disciplinary action

High-risk AI isn’t prohibited, but organisations must demonstrate they have robust data governance, audit trails and traceability, clear documentation of how their AI-led system works, human oversight and the ability to challenge outcomes.

Crucially, a clear red line exists around the adoption of emotion recognition tools. This requirement was already enforced from February 2025, with AI systems designed to infer emotions based on voice tone or “stress analysis”, facial expressions and behavioural or biometric signals made illegal in workplace environments, including contact centres, unless used strictly for medical or safety purposes.

UC Today’s Rob Scott described this as a moment of reckoning for the sector, warning that, “Emotion AI at work is no longer a product category in Europe. It’s a violation of fundamental rights”.

And as of August 2026, vendor assurance is not a defence. Under the Act, responsibility sits with the organisation deploying the system, not the supplier selling it.

Transparency and human oversight are mandatory

The EU AI Act enforces principles many contact centre leaders already support,  but will now do so with the application of legal force, and like GDPR before it, comes with the possibility of fines that run into the tens of millions.

For customer-facing AI, organisations must ensure customers are clearly informed when interacting with AI, that human escalation routes are available, and that AI supports, rather than replaces, accountability.

For workforce-facing AI, leaders must also be able to show:

• AI outputs are reviewable and explainable
• Humans can override or challenge decisions
• No material decisions about people are fully automated

For contact centre leaders, this means AI literacy is now a leadership competency, that procurement decisions carry regulatory consequences and that compliance cannot be retrofitted after deployment. There are some immediate, practical steps to consider:

• Map where AI is used – particularly across QA, performance management and monitoring.
• Query any technology vendor claims around AI and ask directly whether tools perform emotion inference and how EU compliance is handled.
• Separate enablement from evaluation – AI that supports advisors (knowledge, summarisation) carries far less regulatory risk than AI that judges them.
• Establish clear ownership – Whether via governance, an AI lead or executive sponsorship, accountability must be explicit.

Being the first legislation of its kind across the globe, the EU AI Act is groundbreaking. As such, it continues to evolve, however it is very clear in its position that opaque, unaccountable AI has no place in modern workplaces.

For contact centres, this is an opportunity to showcase responsibility around your deployment of AI, and build trust with your colleagues, customers and regulators.