Does your AI setup meet the EU AI Act?

Since 2 August, AI talking to your customers has to say it's AI. By 2 December, it has to prove it too. At Dixa, our AI agents are fully compliant across all channels, including chat, email, WhatsApp and phone.

‍

What is the EU AI Act?

The EU AI Act is the EU's law for artificial intelligence, the first of its kind anywhere. It applies to anyone who builds, sells or uses AI in the EU, including non-EU companies. It sorts AI by risk: some uses are banned, high-risk ones like hiring or credit face strict rules, and chatbots and AI-generated content must be disclosed as AI. The goal is to protect people's safety and rights with one set of rules across the EU. It's being phased in through 2028.

Article 50 of the EU AI Act sets two rules for AI that talks to your customers, whether on chat, email or the phone. First, it has to tell people it's an AI, which has applied since 2 August. Second, what it generates has to carry a hidden mark a machine can detect, which applies from 2 December for AI that is already running. Saying so in a policy somewhere isn't enough.

‍

What does the EU AI Act mean for you?

Your AI has to prove it complies, not just claim it. It's easy to assume your AI meets the bar because your vendor says it does, but a policy page doesn't prove it. 2 December lands in the middle of the Christmas peak, exactly when teams lean hardest on automation, so it's worth knowing where you stand now rather than finding out mid-scramble.

‍

How can Dixa help you comply with the EU AI Act?

Dixa's AI agent, Mim, is already compliant on every channel. We're ahead of the game. Across chat, email, WhatsApp and phone, Mim is built to say it's an AI and to prove it. Teams on Dixa don't have to chase any of it. They spend December on volume, not paperwork.
‍