The businesses that take the lead won’t be the ones with the flashiest UI. They’ll be the ones designing experiences AI can understand and act on. 

Fewer people are starting their journey with a business on a website. Instead, they ask an AI agent to do the work for them. 

Today, that interaction often happens in a search box or bot. Ask a question, get an answer, and move on. But we’re not far from a future where AI agents act on a customers’ behalf and engage directly. 

It may feel like an incremental shift today, but Rob Simms, Customer Service & Support Manager at Linney puts it plainly: “we are now in a world where early adopters have seen proven value from AI, and are exploring the next iteration of improvements and features; while others are playing catch-up.”

As AI becomes the primary way people discover, compare and act on information, the customer experience can’t be designed just for people. It will also have to work for the AI agents acting on their behalf. 

The rise of agent preference

This is where agent preference begins to matter. The question is no longer just “How do we build a great customer experience?”, it’s now “How does our experience work when the visitor is an AI agent?”  

Take travel as an example. Soon a customer might open their personal agent and say “Go book me a flight” or “Find and book the best hotel based on my preferences". Instead of relying on a UI, the agent will converse directly with an airline AI agent to compare options, check availability and complete the booking. Businesses that are discoverable and easy for agents to interact with will win these bookings in the end. 

As Sami Ghoche, Co-founder of Forethought now Vice President of Applied AI at Zendesk puts it: “AI Agents will start to serve other AI agents directly, they’ll be conversing with each other in their own AI language. You don't need a UI for that.”

https://share.vidyard.com/watch/GqzTTF8RHxm7XyFKdsoHRo

This is a major shift in service design. Creating an experience just for the customer who arrives directly isn’t enough. The business has to be understandable and usable by the AI agent representing that customer.

In practice, that could mean:

• Organizing information around real customer intent: structuring your knowledge bases, product details, and support flows so an AI agent can understand what a customer wants to do—not just read flat text on a page.
  
  
• Enabling customer's AI Agents to access actions, safely and easily: Setting up workflows (using tools like Forethought Action Builder) that let AI safely carry out real tasks behind the scenes—like updating an address, checking stock, or processing a refund—without needing a human to click through a visual interface.
  
  
• Providing clear context and guardrails: Giving your AI agents the right rules, policies, and data access so when your customer’s personal agent reaches out, your business’s AI can answer accurately, maintain privacy, and make smooth handoffs to human teams when needed.
  
  

If AI is talking to AI, what does the customer need to consume?

That leads to a more practical question: if AI is engaging behind the scenes,  what does the customer actually need to see?

Not always as much as we think. Some experiences will still require a rich interface, especially where nuance, trust, comparison or complexity matter. But in many cases, customers won’t need to navigate a heavy UI if AI can complete the task directly. The interface may become lighter, more intentional, and more focused on the moments where human attention really adds value. 

The real work may happen behind the scenes, while the customer experiences something simpler, faster, and more seamless.

As Sami Ghoche notes, “a company’s brand website won’t ever be irrelevant, people will always look at things AI doesn’t necessarily need to. Websites may become less about discovery and more about depth.” 

For businesses, that means freeing the UI for humans to explore more clearly. The execution layer lives behind the scenes, while the front end becomes more purpose-built. 

What happens to customer service?

AI will make many things dramatically faster and more efficient. But that doesn’t mean customers will want less human support. In many cases, the opposite will be true.

According to a 2026 Forethought consumer survey, 45% of respondents will give AI only one attempt before escalating to a human. 

This tracks in the real world. As RemoteLock continues to evolve its support operation, the company views Forethought AI agents not as a replacement for people, but as a way to empower teams to handle complex, high-empathy moments while automated agents manage structured tasks behind the scenes.

Because AI raises the baseline for quick, easy resolutions, customer expectations rise with it. They’ll want faster answers, more information, and a quicker path to a human when needed. Websites may become the place customers go when they want to verify, compare, or explore beyond what an AI has already surfaced. The human role will remain essential wherever empathy, judgment, trust, or nuance matter most.

That’s good news for customers: faster service, better routing, more intelligent support, and fewer frustrating handoffs. But it also raises the bar for businesses. “Good enough” digital experiences will no longer be enough.

The future is built for agents

If a customer’s default action is no longer “I’ll search the web,” but “I’ll ask my AI,” then the business experience must be designed for that reality.

The future is not about replacing humans. It is about recognizing the first interaction may no longer be human-to-business. It may be agent-to-business, with the customer one step further back, and the business one step more responsible for being machine-readable, machine-usable, and machine-ready.

Start preparing now. The way people discover and use your business is already beginning to change.

Ready to future proof your customer experience? Learn more here.

‍

‍