In September 2025, OpenAI launched the Instant Checkout feature that allowed users to directly purchase select items from retailers such as Etsy, Walmart and Shopify within the ChatGPT ecosystem. It was touted to be the next big thing in ecommerce, which would enable users to buy directly from OpenAI’s chat interface. Six months later, the company scaled back those ambitions, according to a report from the online media company The Information.
The move followed the lack of adoption among users, who researched for products on ChatGPT but did not use the platform to make purchases. The report, quoting an OpenAI spokesperson, said that the company was now prioritising product search and discovery which would redirect users to buy directly on the retailer’s app or website.
This is broadly where artificial intelligence (AI) is headed in the multibillion dollar global ecommerce market, multiple founders and executives said, citing challenges including changing customer behaviour and not having built enough trust for consumers to buy from the platforms.
In addition, the lack of clarity on whether retailers would have access to first-party customer data, which is relevant information collected from customers such as purchase records, contact details and website use, is also making retailers sceptical, said industry executives.
AI and shopping
There is a massive appetite for AI in shopping. According to a December 2025 report from Morgan Stanley, AI-powered shopping assistants are emerging and could drive $385 billion in US ecommerce spending by 2030. The report highlighted the change in consumer pattern, with 30-40% of the large language model users making purchases.
In India, Google launched shopping in AI mode in search in April, which now includes more than 60 billion updated product listings, with 87% of people reported making more confident decisions using AI mode, the company said in a statement. “In India, shoppers who use AI in their purchase journey interact with 13 touch points on average, compared to just 5.5 for those who do not,” it said.
Indian ecommerce companies such as Flipkart, InMobi and Meesho have invested significantly in AI. Early this year, Flipkart launched conversational AI commerce platform SLAP as a standalone application. Meesho is scaling up its use of AI and recently launched PRISM, an AI-powered platform that analyses user behavioural patterns to recommend products to users. The firm also launched a conversational AI platform, Vaani, early this year.
A senior Indian ecommerce platform executive said that AI will play a significant role in the way people find and discover products, and that the company is investing significantly in this. But end-to-end agentic checkout is a few years away, he said.
AI-focused discovery platforms such as Flash AI are seeing increased conversations. The company has deployed Flash AI with direct-to-consumer platforms and has found that non-AI traffic is converting 6-10% purchases, whereas any query that goes through ChatGPT, Gemini, Flash or Amazon converts at least three times more than the organic traffic. The company is also launching an AI-powered discovery store. However, this is largely restricted to search and discovery, according to founders.
A case in point is the Walmart and OpenAI experiment. In 2025, when OpenAI announced Instant Checkout, the firm partnered with multiple players, including Walmart. According to reports, Walmart found that conversion rates while checking out using ChatGPT were three times lower than the traditional channels, and the firm pulled back to focus on its own shopping assistant Sparky.
Ranjith Boyanapalli, founder of Flash AI, an AI-led discovery platform, said there are two aspects when it comes to shopping and AI—discovery and checkout.
He pointed out that most of the innovation is happening in the discovery part, with checkout still a few years away. “That is one of the reasons why ChatGPT first launched checkouts and rolled them back because they realised that for this to come to pass, you need both customer and merchant adoption,” he told ET.
Neither consumers nor merchants are ready to jump on to agentic checkout right now, he said, adding, “It is not an issue of whether there is enough technology ready; it is a matter of whether the ecosystem, regulators are ready, and of course merchants and consumers.”
Raviteja Dodda, founder of MoEngage, an agentic customer engagement platform that works with ecommerce companies, said multiple things are happening on the agentic commerce side, but the adoption is still low. “When it comes to agentic commerce, it’s important for brands to have first party customer understanding, and are not dependent on a third party. It is not clear yet how that works,” he said.
Zave cofounder Hiren Patel said changing consumer behaviour is a key challenge. “It is tough to change the consumer’s primary discovery destination, (which) is always going to be the commerce platform, and the reason is much deeper than we think,” he said, pointing to the trust that these platforms have built over time with consumers. This partly explains why companies such as Walmart are focusing on their own AI shopping assistants and integrating them into large language models to drive AI-powered product discovery.
Powering the discovery engine
Rashida Bohra, cofounder of ALT Fashion, an AI-led fashion discovery platform, said that just a few years ago she had to explain about AI-led alternative marketing channels. But that has changed as consumer brands understand the need for AI channels. “They are aware that AI is also leading to a lot of commerce, and how it is impacting transactions,” she said.
Brands, Bohra said, are increasingly switching to product catalogues, descriptions and images that are AI readable and more aligned towards AI discovery. “Overall macro environment is very conducive for discovery,” she added.
