Elon Musk backed a bold Grok Bot promise on Wednesday. If the AI agent loses money while running an investor’s bank account, xAI will make the user whole, he says.

One investor is taking him up on it. Yet the numbers tell a different story. It starts with a $100 liability cap sitting in xAI’s own terms.

Inside the Grok Bot Bank Account Experiment

Teslaconomics, a Tesla and xAI investor, asked a simple question on Wednesday. Had anyone connected Grok Bot to a bank account? In a post on X (Twitter), he said the agent could track spending, pay bills, and catch strange charges.

He even wondered whether he still needed a personal banker. His partner reportedly told him an outright no. She fears what an AI with that much access could do.

However, Elon Musk is confident, and even guaranteed reimbursement should the Grok bot mess up.

Follow us on X to get the latest news as it happens

Grok Bot launched in beta on August 11. According to xAI, each agent runs day and night on its own cloud computer. It signs into websites like a human and keeps working while its owner sleeps.

The experiment also fits Musk’s bigger money push. X Money went live with peer-to-peer payments in June.

A $100 Liability Cap Sits Behind the Grok Bot Promise

While Musk’s pledge sounds like insurance, his company’s paperwork says otherwise. xAI’s consumer terms offer outputs and agentic actions on an as-is basis.

Those terms cap most claims at the fees a user paid, or $100, whichever is greater. Bot access comes with a $30 monthly SuperGrok plan, so a year of fees totals $360. That is pocket change next to a drained checking account.

A reply on X (Twitter) does not rewrite that contract. Unless xAI puts the guarantee in writing, any refund rests on Musk’s goodwill.

US banking rules add a sharper edge. Regulation E, the federal rule covering electronic transfers, protects customers from unauthorized payments. However, a transfer loses that label when the customer handed over account access.

Give a bot your login, and standard fraud protections may not apply.

History gives skeptics more ammunition. In May, hidden instructions inside a malicious NFT tricked an AI into moving money, a technique called prompt injection. The attack drained roughly $150,000 from a Grok-linked Bankr wallet. About 80% of the funds later came back.

Weeks later, a separate breach hit 14 user wallets on the same platform, and Bankr pledged full reimbursement. No verified case of Grok Bot mishandling a real bank account has surfaced so far.

Even so, prompt injection could prove costlier when the target holds live bank logins instead of a crypto wallet.

The Grok bot experiment may therefore double as marketing for a product xAI wants inside daily money management. The first real error will show what Musk’s words are worth.

The answer may be a quiet refund, or a very public test of that $100 cap.

Source: BeInCrypto