Meta’s AI assistant Muse has triggered a digital tug-of-war as Canadian banks block third-party bots from client accounts while retailers embrace them. Major financial institutions warn that sharing credentials with AI agents jeopardizes fraud protections, setting up a sharp clash between consumer tech innovation and traditional banking security.
Canadian Banks Block Personal AI Agents Over Security and Fraud Risks
Some of Canada’s largest financial institutions have moved to block AI agents like Meta’s Muse assistant from directly accessing online bank accounts and other authenticated digital properties. The restrictions arrive as the financial sector confronts automated software that can monitor balances, track transaction histories, and potentially execute purchases on behalf of users.
Royal Bank of Canada and Toronto-Dominion Bank confirmed actions to restrict third-party agent connections. TD stated that it blocked all third-party personal AI agents from direct access to banking information requiring authentication while appropriate safeguards and oversight are established.
Clients should never share their TD log-in credentials, passwords, PINs or one-time passcodes with an AI agent or anyone else.
Ada Lovelace Rampersaud, TD spokesperson
Similarly, RBC informed employees via a company memo that it disabled third-party personal AI agent access to authenticated digital properties. Canadian Imperial Bank of Commerce stated that it does not support personal AI agents directly opening or accessing accounts. Bank of Nova Scotia is limiting third-party AI agents, while Bank of Montreal stated that it does not permit AI agents to access its digital properties or execute transactions. RBC spokesperson Jeff Lanthier noted that the institution carefully weighed the risk of third-party personal AI agents before deciding to block these applications from its online banking platforms. CIBC spokesperson Deborah Rowe confirmed that her bank is following developments in personal AI agents closely as these technologies evolve.
Retail Integration and Consumer Shopping Friction With Meta Muse
Major e-commerce players are welcoming automated assistants while banks slam the door. Meta confirmed that the Muse app had been downloaded more than ten million times across the United States after leaping to the top of the iOS App Store’s list of free apps following its launch early last month.
Retail consultants note a fundamental struggle over customer ownership.
There’s a real tug-of-war going on right now over who owns the customer experience.
Uwe Stueckmann, co-founder of retail consultancy Innovative Marketing
Real-world testing reveals that automated shopping remains imperfect. CBC News learned from a Toronto resident that Muse arranged a Marketplace transaction without his awareness, even disclosing his home address to an unknown person.
BMO CEO Darryl White Warns Canadians on Password Sharing and Fraud Protections
Speaking at the Toronto Global Forum, Bank of Montreal CEO Darryl White cautioned that decades of online shopping have conditioned people to surrender sensitive data too easily. White urged Canadians to withhold banking usernames, passwords, credit card numbers, social insurance numbers, and PINs from AI assistants.
I think we’ve got a ways to go on educating people.
Darryl White, BMO CEO
Handing over login credentials to an outside application introduces severe financial risks. Sharing credentials with an autonomous software agent can jeopardize those fraud protections, leaving consumers liable for losses. Aiming to establish more secure rules, the Consumer-Driven Banking Act achieved royal assent in March 2026, alongside proposed regulations released in the Canada Gazette that would permit Canadians to share their financial data with certified institutions supervised by the Bank of Canada.
International Security Crackdowns Across Global Financial Institutions
The defensive posture extends far beyond Canada. Major U.S. banks including Morgan Stanley, Bank of America, JPMorgan Chase, and Wells Fargo updated terms of service to prevent third-party agents from taking autonomous actions inside customer accounts. Financial executives emphasize the extreme hazards of unverified software interacting with core monetary infrastructure. Citigroup also took steps to limit third-party AI agents, stating that it does not permit outside agents or assistants to log into customer accounts or transact directly within its digital banking platforms, though it allows limited read-only access through approved open-banking providers.
We’re not going to allow just any… outside agent to come in inside the company, because it’s just too risky for clients and for JPMorgan.
Jamie Dimon, JPMorgan Chase CEO
Financial institutions are steering developers toward safer alternatives, such as limited, read-only data sharing through approved open-banking intermediaries like Plaid. These configurations allow budgeting apps or assistants to view financial summaries without gaining the ability to transfer funds or pay bills independently. Brian Moynihan noted that Bank of America is restricting AI agents’ access to consumer deposits and working with companies to develop protocols on agents, while Wells Fargo updated its online access agreement to establish that authorized representative and agent transactions are considered binding unless the user notifies the bank of unauthorized activity.
How will regulators reconcile consumer-driven banking laws with the absolute blocks banks place on third-party credentials?