🔍 Read the full analysis: A Clear Guide To AI Agents, Bank Passwords, And Money Safety on ThorstenMeyerAI.com
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Bank of America has warned customers not to share bank login details with AI assistants and says customers are responsible for activity resulting from that access. Visa has identified trust as a barrier to agent payments, while Robinhood says customers have placed more than $250 million in accounts used for AI-agent trading. The legal treatment of agent-initiated transactions remains unsettled, and narrower permissions may reduce exposure.
Bank of America has warned customers not to share their usernames and passwords with AI assistants, saying they are responsible for activity resulting from access they provide. The warning comes as Visa identifies trust as a barrier to agent-led payments and Robinhood says customers have put more than $250 million into accounts used by AI agents, highlighting unresolved questions about control and liability.
The email, titled “Protecting your finances in the AI World,” says login credentials should not be shared with another person or an outside service, including an AI-enabled tool or agent acting on a customer’s behalf. A customer posted a screenshot on Bluesky; TechCrunch reported on it on October 27, and a Bank of America spokesperson confirmed the email was authentic. The bank also said it is excited about what AI can deliver when used responsibly.
The warning is not a claim that every AI agent is unsafe. It reflects a basic access problem: an agent using a customer’s password may be able to act with the same account permissions as the customer, while the bank may have difficulty distinguishing that activity from the customer’s own actions or from a stolen password. The source account says Visa told investors this week that trust is the biggest barrier to AI agents paying for goods and services. Robinhood said customers had already placed more than $250 million into accounts that AI agents trade; the source does not specify the measurement period or a comparison baseline.
Bank of America’s email states that customers who share credentials are “responsible for activity resulting from that access.” That is the bank’s warning, not a court ruling on how responsibility would apply in every dispute. The source also describes Robinhood as using separate accounts and spending limits, while quoting the company as saying customers “assume all risk” for trades by AI agents. Those safeguards and terms are distinct: limits may constrain access, but do not by themselves settle who bears losses.
Before your agent touches money
Give it a valet key, not the master key
Logging in with your password makes the agent you. A credential made for the agent can do only the job you gave it, and you can take it back.
- What it opensEverything you can do: transfers, payees, settings, statements
- LimitsNone that the agent cannot get around by logging in as you
- What the bank seesYou, or a thief. It cannot tell which
- If a web page tricks itThe trick reaches your whole account
- To take it backChange your password everywhere it was used
- If money goes missingUS rules can exclude transfers by someone you gave access to; Bank of America: “you’re responsible”
- What it opensOnly what you allowed: read-only, one card, one pot of money
- LimitsBuilt in: amount, merchant, time window, approval on your phone
- What the bank seesAn identified agent acting for an identified person
- If a web page tricks itThe damage stops at the limit
- To take it backDisconnect the agent or delete its card. Your password never moved
- If money goes missingDepends on the terms. Read them first: some still put all risk on you
A comparison of two set-ups, not a measurement. Which limits exist depends on your bank, card and app.
Sources: Bank of America customer email, confirmed to TechCrunch (27 Oct 2026); 12 CFR 1005.2(m); OWASP LLM01. Comparison: Thorsten Meyer.
Powered by Thorsten Meyer AI
The Cost of Giving Agents Access
For customers, the distinction between a stranger stealing a password and an agent they connected acting unexpectedly could affect whether a transaction is treated as unauthorized. The source material points to uncertainty in US and EU rules, but does not establish how regulators or courts would decide a particular AI-agent case. Customers should not assume that familiar fraud protections automatically apply after they provide an agent with account access.
There is also a security concern beyond accidental spending. Agents may read websites, messages, invoices and files as they carry out tasks. The source describes prompt injection as instructions hidden in those materials that can steer an AI system. If an agent has broad banking access, a misleading or malicious instruction could have more serious consequences than if it can only read account information or use a tightly limited payment method. The practical issue is therefore not simply whether to use AI, but what authority the tool receives and what review is required before money moves.
Products Worth Considering
Bose QuietComfort Wireless Bluetooth Headphones with Active Noise Cancellation, Mic, USB‑C Charging, Deep Bass, 24‑Hour Battery – White Smok
Amazon Kindle Paperwhite Signature Edition (12th Generation, 2024), 32 GB, Metallic Black
DJI Osmo Pocket 3 Creator Combo Vlogging Camera with Mic 2, 1 CMOS, 4K/120fps, 3-Axis Stabilization, 128GB Card
As an affiliate, we earn on qualifying purchases.
Banks Face Agent Activity
The warning follows a report by Axios on October 22, 2026, which cited unnamed sources who said major US banks were finding third-party AI agents inside customer accounts. According to that report, the agents were generally logged in using customers’ own passwords, and banks often could not tell whether the activity had been directed by the account holder. The account provided here does not independently verify those reported cases.
Financial services firms are beginning to address agent payments, but their approaches differ. The source says Robinhood has used separate accounts and spending limits for AI-agent trading, alongside terms assigning customers risk. For other use cases, narrower forms of access could include read-only permissions, approval before each payment, limited virtual cards, or a token restricted by merchant, amount and time. These are presented as safer design patterns in the source, not as protections guaranteed to be available at every bank or platform.
““Your username and password should never be shared with another person or provided to an outside service, including an AI-enabled tool or agent that offers to act on your behalf.””
— Bank of America, in its customer email
Liability Rules Are Untested
No regulator or court ruling identified in the source material determines whether giving an AI agent a banking password counts as furnishing an account access device under US Regulation E. The source author’s interpretation is that the rule could apply, but explicitly says it has not been tested in court. Likewise, whether sharing credentials with an agent would amount to gross negligence under the EU Payment Services Directive has not been determined in court, according to the source.
It is also unclear how often agents have initiated disputed transactions, how banks distinguish customer-directed activity from unauthorized use, and what recourse customers would have in specific cases. The source’s account of banks finding agents in customer accounts relies on Axios reporting citing unnamed sources. The $250 million Robinhood figure is not accompanied by a time period or baseline, so it cannot show a rate of growth or how many customers or trades were involved.
Check Permissions Before Connecting
Customers considering an AI financial tool can check whether it requests a password or offers limited, revocable access. Where available, read-only access, payment-by-payment approval, a capped virtual card, or a separate account with only a limited balance can reduce what an agent can do. Users should also read the service’s terms to see who bears responsibility for transactions and whether they can disable access promptly.
The next developments to watch are how banks, payment networks and trading platforms define agent permissions, handle disputed activity and communicate customer protections. Any legal clarity will depend on regulators, courts and the specific facts of a transaction. Until then, Bank of America’s warning is clear: do not treat a bank password as a small, task-specific permission.
Key Questions
What did Bank of America tell customers?
It warned customers not to share their usernames or passwords with outside services, including AI tools and agents, and said customers are responsible for activity resulting from access they provide.
Does the warning mean AI agents cannot be used for financial tasks?
No. The warning concerns sharing full banking credentials. The source describes narrower options such as read-only access, payment approvals, capped virtual cards and accounts kept separate from a customer’s main funds.
Will a bank reimburse money spent by an AI agent?
That is not settled by the information available here. The source says the application of US and EU rules to AI-agent access has not been tested in court, and outcomes may depend on the facts and the applicable terms.
What is prompt injection?
It is an attempt to steer an AI system by placing instructions in material it reads, such as a webpage, email or file. An agent with broad financial permissions could pose a greater risk if it follows harmful instructions.
What should customers check before connecting an agent?
Check what access it receives, whether each payment needs approval, whether limits can be set, how access can be revoked, and what the provider’s terms say about responsibility for losses.
Source: ThorstenMeyerAI.com
