3 Ways To Escape Chatbot Backlash And Design AI Interfaces People Want

Forbes

23-07-2025

Chatbots continue to multiply despite consumers expressing fatigue and user experience experts say ... More there are multiple considerations when picking an interface for AI.
Many people immediately have a negative reaction when a chatbot automatically pops up and offers to help. Chatbots have increasingly become the face of AI for the general public, but they are not always the best way to leverage AI to help users. The rate of chatbots has been growing, and their compound annual growth rate financially is estimated to grow 23.3% from 2025 to 2030 according to Grand View Research. Companies eagerly embraced implementing chatbots as a way to solve customer problems, but their overuse and sometimes underbaked implementation has amplified negative opinions of AI. Companies building AI powered products should consider if a chatbot is the right user interface to help the user.
Negative Perception of Chatbots
In a Tidio survey of 1015 people, 82% of consumers said they would be willing to talk to a chatbot if they had to wait to get ahold of a human representative in customer service contexts. However, in real life, many consumers feel there is a lot lacking. AI can even hurt emotional trust in a solution, or even trigger fear according to a 2024 study by Mesut Cicek and Dogan Gursoy.
'AI pops in and starts writing for me or asking my questions and how can I help you? And I'm like, I don't need any help…stop hijacking my messages,' Moira Morton, a sales and marketing consultant, shared with me in a conversation. Morton said often the chatbots pop up without her engaging with them at inappropriate times like booking travel or in software she uses for work. It creates a frustrating experience that she finds hard to escape from. When this happens, she says she feels like saying 'don't be sorry…just go away.'
Building the Right Interface for the Right Problems
'When companies replicate the chatbot without understanding its underlying [capabilities]
'Too many teams begin with 'we can use AI' instead of asking why. The result is often a solution in search of a problem. The most successful AI experiences are grounded in a clear understanding of user needs and business goals,' Chaturvedi said. Companies need to be honest with themselves if they are building the right solution or if they are checking off a checkbox saying they are 'doing AI.'
The Right Approach
For many companies, building an AI experience is new to them. Chatbots seemed like an easy way to get started. There are multiple things they need to consider.
What Problem Is Being Solved
Companies should be honest with themselves about what problem they are actually trying to solve. Not all problems need to be solved with AI. And the problems that should be solved with AI shouldn't necessarily be solved with a chatbot. The interface should align with the best way to solve the problem based on the value you are trying to create for the customer.
Understand Your User
Understanding your user has several layers. The first layer is context around how your users solve their problem today. Consider what system or interface they are normally in when they solve this problem. Think about how they feel about their current solution. It's possible they don't see an issue with how they solve the problem today, so your solution needs to be instantly obvious to them about why it's better.
You also need to consider the visual context the user needs to get value from the product. While the simplicity of the text box that can create almost anything seems like a great choice, for many users, this is the equivalent of handing them a blank page. There is a reason most web pages aren't blank pages: because many users need additional context on the page to understand what is possible and how to get the value they need quickly.
Trust is also important to consider in any interface, especially AI-powered ones. Understanding what type of information your user needs to be able to trust what you share with them is required. If they don't trust what you are showing them, regardless if it's in a chatbot or a more integrated experience, your product will struggle to get adoption.
Test and Derisk
AI can be expensive to get wrong. Before going too far down any one path, companies should test the output and potential interface for the new AI feature. With AI output, companies should be sure the information they are sharing with the user actually provides real value. They also should share design or prototypes with users to get early feedback on where there is still friction in the flow, even where AI will be a part of the solution. There can be a tendency to over emphasize a solution that leverages AI, but some of the most AI-laden products like Instagram or TikTok are not thought of as AI products, even though AI is in so many facets. Sometimes the best AI products are so seamless, no one thinks about them as AI.
AI can help solve a lot of problems, chatbots shouldn't be the assumed answer to every problem. Companies implementing AI should have a solid understanding of the problem they are solving and make sure they solve it in the right way in the right context for their users. Rushing to implement a chatbot could hurt the trust in your product and early AI features. Take the time to choose the right interface with the right level of accuracy and detail for your audience to avoid hearing users like Morton wish your AI solution would just 'go away.'