As artificial intelligence rapidly integrates into ecommerce, many retailers are scrambling to release AI-driven shopping assistants, chatbots and automated purchasing. However, new data from Omnisend reveals that consumers aren’t ready to hand over control of their wallets just yet.

Omnisend’s survey found that 66% of consumers would refuse to let AI make purchases on their behalf even if they get a better deal, despite recognizing AI’s benefits in shopping. When asked which current AI-powered features they found most useful, 38% of shoppers cited personalized product recommendations, while 31% said AI helps them discover products faster. However, while consumers appreciate AI’s ability to enhance the browsing experience, they draw the line when it comes to AI making final purchasing decisions for them.

“Consumers are open to AI enhancing their shopping experience, but there’s a big difference between receiving personalized recommendations and handing over full purchasing control,” said Greg Zakowicz, senior ecommerce expert at Omnisend. “Large-scale adoption of AI-driven purchasing requires a fundamental shift in consumer behavior — and that’s not happening anytime soon. Shoppers want to know exactly what they’re buying, whether it’s a size, color or brand. Trusting an AI to make those decisions autonomously is a tough sell.”

While AI holds the promise of making shopping smarter, many consumers remain skeptical about its true intent and effectiveness. According to the survey, 42% of shoppers say that a flurry of targeted ads is the main drawback of ecommerce AI in its current state. This reinforces a perception that AI serves the retailer’s bottom line more than customers’ needs.

[Read more: Generation next: Capturing millennial, Gen Z shoppers]

A further 58% of consumers say they are worried about how AI handles their personal data. Transparency remains a major issue, with 28% saying they don’t trust any company with their data handling.

Beyond trust concerns, consumers aren’t convinced AI is “smart” enough to handle key parts of the shopping experience. Thirty nine percent of shoppers have abandoned purchases due to frustrating AI interactions, such as inaccurate recommendations or poor chatbot experiences. Meanwhile, nearly half (48%) say AI should focus on improving customer service rather than making automated purchase decisions.

“AI is a tool that’s only as good as the problems it solves,” Zakowicz said. “As companies race to integrate AI into their customers’ shopping experiences, they need to be more focused on its problem-solving qualities versus its potential. Providing poor experiences, whether by AI or a human, will result in lost sales.” 

[Read more: Forecasting the future]

Implications for retailers:

As AI becomes more pervasive in the retail space, the challenge for brands isn’t just about adopting new features but about adopting them responsibly. The data suggests there is a tremendous opportunity and significant risk in how retailers deploy AI.

Here’s what the findings mean for brands moving forward, based on insights from Zakowicz:

• Trust-building must be a priority. Retailers must be transparent about data usage and give customers granular control of their information. Integrating AI in a way that genuinely benefits shoppers and not just boosts sales reinforces that trust.
• Personalization vs. intrusiveness. Relevant recommendations can improve the shopping experience, but excessive retargeting pushes consumers away. Providing shoppers control over how and when they receive suggestions reduces the “creepiness” factor.
• The need for human-centered AI. Chatbots can handle common questions efficiently but must escalate complex issues to real people. Solutions based on actual customer needs come across as supportive rather than purely sales-driven.
• Consumers demand reliability. Repeated AI errors quickly erode trust. Retailers that continuously refine their AI tools and address user feedback can significantly boost performance. Rigorous QA and an iterative approach can help build confidence and long-term loyalty.