Your AI Now Has to Introduce Itself
Disclosure rules are live, shoppers are checking the work, and CX teams need proof customers can actually see.
Your daily signal on AI and CX — minus the hype.
DCX Stat of the day: 33% of UK shoppers worry AI recommendations may not show them the best products because the answers could be inaccurate or commercially influenced. TechRadar
In this issue:
→ The AI disclosure rule is now live
→ A label needs evidence behind it
→ Useful recommendations still lose shoppers at checkout
→ A no-code builder puts a human face on the agent
→ A regulator-built checker maps the obligations
→ Your disclosure needs an owner
🔍 DEEP DIVE
The Disclaimer Just Became a Journey Step
As of August 2, Article 50 of the EU AI Act applies to customer-facing AI interactions and certain AI-generated content. Providers must make sure people know when they’re interacting directly with AI, unless that’s already obvious. The notice has to be clear, distinguishable, accessible, and present by the first interaction or exposure.
That puts a small but consequential moment inside the customer journey. The chatbot greeting, voice-agent opening, synthetic spokesperson, emotion-recognition notice, or AI-generated public-interest content label can’t sit buried in a privacy policy nobody reads.
Here’s the practical problem: adding the words “AI assistant” is easy. Proving the notice appears across every language, device, channel, vendor update, and redesigned flow is harder. Customers need enough context to calibrate trust. The business needs evidence that the disclosure shipped, stayed visible, and has a named owner when the experience changes.
Bottom Line: Transparency is now a product requirement with an operating trail, not a sentence Legal hands to Design.
📬 Copy-Paste Take
If an AI system talks to customers, the disclosure belongs in the tested journey. Name where it appears, who owns it, how accessibility is checked, and what evidence proves it survived the latest release.
🧭 OPERATOR PLAYBOOK
Run the Monday-Morning Disclosure Test
Pick one live customer-facing AI experience. Open it like a customer would, before anyone explains how it’s supposed to work.
Audit every disclosure moment for four things:
Timing: Does the customer see or hear it by the first interaction?
Clarity: Does it say AI plainly, without making the customer decode a brand name?
Accessibility: Does it work across language, screen reader, voice, mobile, and low-vision needs?
Evidence: Can the team show a current screenshot, recording, release check, and owner?
Then test whether a vendor update, channel handoff, or translated flow makes the notice disappear.
Ask your team: If this disclosure failed tomorrow, who would know first and who could fix it?
Signal: A disclosure you can’t test is a disclosure you don’t control.
📊 MARKET REALITY CHECK
Useful Isn’t the Same as Trusted
AI is already shaping retail discovery. Twenty percent of UK consumers now start product discovery with AI search platforms, ahead of Facebook, TikTok, or Instagram individually. And 87% say AI recommendations are useful.
But only 55% have bought after receiving one. Just 10% say they’d spend more time with an AI shopping assistant than a human, while 65% still prefer a physical store for high-value purchases. The data doesn’t prove disclosure alone closes that gap. It does show that usefulness hasn’t removed the need for verification, comparison, and human confidence before money moves.
Why it matters: AI can earn a place in discovery without earning the customer’s decision. CX leaders still own the proof, choice, and recovery that turn an answer into a purchase.
Useful recommendation + visible proof + customer control = a chance at conversion
🧰 TOOL WORTH KNOWING
Tavus PAL Maker
What it does: Tavus PAL Maker lets teams describe a personified AI agent in plain language, build it without code, and deploy it as a hosted page, inside a site or app, or into Google Meet. Teams can add a knowledge base, memory, goals, guardrails, a custom face, and a custom voice.
CX use case: Prototype an onboarding guide, customer expert, support agent, patient-intake assistant, or sales concierge that can see, hear, respond, and bring visual information into the conversation.
Worth watching because: It moves customer-facing AI beyond a chat box and toward a presence that can feel like a person. That raises the bar for disclosure, consent, knowledge boundaries, memory controls, escalation, and recovery.
Bottom line: PAL Maker lowers the cost of putting a face on an agent. It does not lower the burden of deciding what that agent can know, remember, promise, and do.
The DCX AI Today - AI Tool Directory - If you lead a CX team and want a curated shortlist of tools worth evaluating, this is your starting point.
📡 90-SECOND CX RADAR
EU AI Act Compliance Checker: The European Commission’s official beta asks structured questions about an AI system and maps possible obligations for providers, deployers, and other operators. Use it to start a cross-functional scoping conversation, not as a certification or legal determination.
✅ YOUR MOVE
AI transparency has moved from guidance into the live customer journey.
Retail behavior shows why that matters: customers can find an AI answer useful and still stop short of trusting it with the decision.
This week, test one chatbot, voice agent, recommender, or AI-assisted content flow as a customer. Capture the first interaction. Check the language and accessibility. Confirm the owner. Then add that evidence to the release gate so the disclosure doesn’t vanish the next time the experience changes.
If the customer has to hunt for the AI disclosure, the journey has already hidden the part they needed to know.
Until tomorrow,
👥 Share This Issue
Think of one person who’s wrestling with AI in CX right now
and forward this to them.
I’m obsessed with Wispr Flow Pro! Get a Free Month on me.
If someone forwarded this to you, they thought you needed to see it before your next AI planning meeting. Get your own copy.