Musk Says SpaceX Will Reimburse Users for Grok Purchase Mistakes

Elon Musk said SpaceX will reimburse users if its Grok AI agent makes a purchasing mistake or exposes their card details. In a Thursday post on X, Musk wrote that the company would make users whole, responding to a user who said Grok booked a flight after being sent a photo of a credit card. The user said the airline’s website was confusing but that Grok completed the transaction and sent a booking confirmation. Musk described the episode as working well, adding that leaks had not yet happened. The statement did not clarify the scope of the pledge, including whether it would cover unauthorized purchases, incorrect bookings, fraud by third-party merchants or consequential losses such as missed travel. SpaceX did not immediately respond to a request for comment. The promise stands out because AI agent providers generally place responsibility on users. Meta’s Muse requires approval before purchases, and Grok’s terms mention interactions with financial institutions, but none offers such an explicit reimbursement commitment. The remarks come as AI agents increasingly access payment methods, personal accounts and online services, exposing new operational risks. In one recent Grok-related case, Shane Mac, chief executive of software company XMTP Labs, said a personal-finance agent he created posted his bank balances and spending details in his company’s executive Slack channel after confusing two similarly named chats. Grok’s team later adjusted its permission controls. Other agent users reported similar errors, including an Instinct user whose agent ordered 20 pounds of chicken wings during a grocery shopping task and a Muse user whose agent accepted a low Facebook Marketplace offer for a keyboard and shared his home address with the buyer. The Muse user said he had selected an automatic approval setting but believed the agent would still seek confirmation. Musk’s pledge adds little detail, but it highlights a growing question for companies deploying autonomous financial agents: whether users will continue to bear most of the risk or whether providers will begin to share responsibility when their systems make costly mistakes.