Next year, people using ChatGPT can make purchases on the platform through PayPal, which is bringing what the company calls “tens of millions of merchants” to OpenAI’s popular chatbot.
The deal between PayPal and OpenAI was first reported by CNBC, which said that PayPal will be enabled on ChatGPT sometime in 2026 and that merchants will be able to sell their inventories directly on ChatGPT.
Don’t miss any of our unbiased tech content and lab-based reviews. Add CNET as a preferred Google source.
In a video from PayPal showing how PayPal integration might work in ChatGPT, a customer asks the chatbot for advice on running shoes, writing: “Help me buy running shoes for Maya, size 7.5. She likes bold colors and city runs. Budget $120.”
The prompt brings up a suggestion for Reebok running shoes and buttons to buy the shoes with PayPal or “Pay another way.”
The deal makes PayPal one of the first e-commerce services to be incorporated with OpenAI’s service and the first payment wallet announced for ChatGPT, which has about 800 million weekly users. OpenAI recently made deals with Walmart, Shopify and Etsy as well.
PayPal CEO Alex Chriss told CNBC: “We’ve got hundreds of millions of loyal PayPal wallet holders who now will be able to click the ‘Buy with PayPal button’ on ChatGPT and have a safe and secure checkout experience.”
Adding PayPal to ChatGPT has been a requested feature on OpenAI’s developer community forum.
(Disclosure: Ziff Davis, CNET’s parent company, in April filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.)
How secure will chatbots be for shopping?
Given the millions of possible shoppers using AI assistants like ChatGPT, the potential for e-commerce to shift from the web to AI platforms is enormous. It could allow recommendations for buying, more in-depth product comparisons, AI tools that are better at remembering customer preferences and smoother checkout processes.
“Fewer abandoned carts, richer customer interactions and rising consumer expectations for instant, personalized experiences,” said John Paul Cunningham, chief information security officer at Silverfort, a Texas-based identity security company.
But Cunningham warned that people who use chatbots for shopping may be vulnerable if they’ve given these tools too many permissions, such as access to their financial data, that isn’t properly secured on their end.
“Chatbots add another layer of risk: prompt-injection attacks that trick them into revealing data or pushing fake product links to phishing sites,” he said.
If a chatbot service were to be the victim of a data breach, it could expose personal data just like any other online service with access to that information.
Add to that emerging AI agents that are constantly improving and could make autonomous decisions that are difficult to keep secure.
“We may need to create super-policing AI agents to monitor, control, and secure the e-commerce experience in real time, paired with better identity verification tools to ensure transactions stay tied to verified individuals,” Cunningham said.
The good news is that larger merchant providers like PayPal have dealt with fraud and security risks for decades. PayPal may be better positioned than smaller merchant networks to combat potential fraud on AI platforms.
“Established platforms like PayPal significantly mitigate fraud risks today,” he said. “PayPal certainly has the maturity and funding to apply toward securing such a platform.”
That said, a payment provider can only do so much if AI tools have vulnerabilities outside its control.
“There may be fundamental flaws in the underpinnings of AI and how it works that may prove difficult to remedy,” Cunningham said.
