the voice of customer experience technology
Front page
NewsJourney Orchestration2h · 15:01 BST · 4 min read

AI Agents Are Becoming the New Customer Front Door

Meta, Sierra, Stripe, Shopify, and Walmart are setting rules for personal AI agents, while Adobe, Hightouch, and Smart Communications tackle the systems behind the journey. The challenge is no longer just AI adoption, but deciding who controls the customer relationship.

AI agents shaping customer engagement journeys across Meta, Adobe, Shopify, Stripe

A customer arrives on your website, asks a question, checks availability, changes an order, and completes a purchase.

This seems to be a completely standard e-commerce interaction, except the customer may never have visited.... their AI agent did.

That possibility is moving quickly from a future-facing thought experiment into product roadmaps, technical standards, and customer engagement strategy.

Businesses have spent years refining digital journeys for people. Now, they must prepare for software acting on a person’s behalf, as Kevin Miller, Head of Payments at Stripe, explained:

“A customer relationship doesn’t start or end at checkout. When customers send an agent, they expect the same service they’d get themselves.”

That is the central challenge for this week’s customer engagement news. AI agents may remove steps from a journey, but they also introduce new questions around identity, authorization, data access, and who gets to shape the experience.

Editor’s Pick: Building Rules for the Agent-to-Business Journey

Meta and Sierra have announced the Personal Agent Protocol, an open standard developed with partners including Genesys, Instinct, Rocket, Shopify, Stripe, and Walmart.

The protocol aims to give companies a clearer way to recognize when an AI agent is acting on behalf of a customer and define what that agent can do.

It covers authentication, visibility, and consent, with customers deciding whether an agent receives read-only or write access.

That is important because personal agents are increasingly expected to handle everyday tasks, from booking flights and scheduling appointments to making purchases and managing returns.

Today, many agents attempt these tasks by navigating a company website as a person would. That is slow, unreliable, and leaves brands uncertain whether the traffic comes from a legitimate customer representative or an unauthorized bot.

Sierra’s proposal gives businesses several options. They can allow agents to use their website, connect through APIs, or hand work off to a company-owned agent for more complex interactions.

The protocol is still in the early stages, with Sierra planning to publish its v0.1 specification later this month. Yet it is a meaningful sign that the conversation is moving beyond agent demos and into the rules that will govern agent-led customer journeys.

The crucial principle is that customers remain in control. If that gets lost, the protocol risks becoming another layer that gives platforms and brands more visibility while leaving consumers with less.

Adobe Takes Agentic AI to Heathrow’s Passenger Journey

Adobe has expanded its partnership with Heathrow Airport to transform the digital experience for millions of passengers using AI and customer experience technology.

Heathrow’s partnership with Adobe is intended to bring more personalized digital experiences to a journey that begins well before the airport and continues after landing. That makes it a stronger test of AI-powered orchestration than many retail examples.

Travel companies have long held large volumes of customer data. The challenge is whether they can use it at the right moment without making a passenger feel monitored or overwhelmed.

If Adobe can help Heathrow make a complicated physical journey feel less stressful, it will be a useful proof point for agentic customer experience.

However, if it becomes another digital layer sitting on top of fragmented operations, passengers will be less impressed.

Hightouch Buys the Data Plumbing Behind Marketing Agents

Hightouch has acquired Strata, an agent-driven platform that helps teams build and maintain custom API integrations without code.

Hightouch’s platform helps marketing and data teams activate customer information across the wider technology stack. Strata adds a way to create and maintain the integrations needed to connect those systems.

The promise is that internal marketing agents gain access to a broader, more current set of signals before deciding which audience to target, which message to send, or which customer journey to trigger.

That should improve relevance. It could also magnify errors if data quality and permissions are not tightly controlled.

The customer engagement market is full of vendors promising autonomous execution. Hightouch’s acquisition is a reminder that dependable automation still relies on the unglamorous work of connecting systems properly.

Smart Communications Focuses on the Legacy Problem

Smart Communications has expanded its AI-powered migration offering for enterprises attempting to modernize legacy customer communications systems.

It is an important counterpoint to the week’s agent-focused announcements. Many organizations cannot move from static communications to agentic journeys simply by adding an AI layer to the top of an aging stack.

Customer communications are often buried in legacy templates, disconnected approval processes, and compliance workflows.

Modernizing those foundations is less eye-catching than launching an AI agent, but it may matter more to customers receiving confusing, inconsistent, or outdated messages.

The real question is whether enterprises can modernize quickly enough without breaking the communications they already rely on.

Look Ahead

The next customer engagement battleground may not be a website, app, or messaging channel. It may be the interface where a customer’s AI agent meets a brand’s systems.

The businesses that win will not simply be those with the most capable agents. They will be the ones that give customers clear control, keep their data accurate, and make an agent-led interaction feel as dependable as a human one.

rate this story
helps rank stories across CX Today
The discussion0 takes · attributed & checked

Does this reflect your experience?

opening the room…
Read nextordered by techtelligence · every pick explained
same beat · Journey Orchestration

Why Churn is a Customer Experience Warning, Not a Metric

12 Oct 2026
same beat · Journey OrchestrationRetailers Cannot Allow WhatsApp to Become a Spam Channel9 Oct 2026same beat · Journey OrchestrationAI Handoffs Should Feel Seamless. Why Aren’t They?8 Oct 2026