SKN CBBA - ...
SKN CBBA
Cross Border Banking Advisors
SKN | Major US Banks Restrict Third-Party AI Agents as Liability Risks Shift to Customers

Finance

SKN | Major US Banks Restrict Third-Party AI Agents as Liability Risks Shift to Customers

By Or Sushan

•

November 22, 2026

Key Points

  • Wells Fargo’s updated agreement warns customers that AI agents are used at their own risk, while several major US banks have introduced restrictions or liability warnings for third-party AI tools.
  • JPMorgan Chase, Bank of America, Citi, PNC, and Citizens have adopted measures addressing AI agents’ access to customer accounts and the responsibility for transactions they initiate.
  • Banks are simultaneously restricting external agents while expanding their own AI-powered tools, highlighting the growing importance of security, accountability, and control in AI-driven financial services.

US Banks Tighten Restrictions on Third-Party AI Agents

Wells Fargo is introducing updated customer terms addressing the use of artificial intelligence agents, with the agreement taking effect November 24. The language places responsibility on customers for using external AI agents to interact with their financial accounts, stating that such activity is undertaken at their own risk.

The policy reflects a broader shift across the US banking sector as consumer AI agents increasingly gain the ability to navigate websites, access accounts, and execute transactions on behalf of users. Although agent-driven transactions currently represent a limited share of online commerce, banks are moving to establish clearer boundaries around access, authorization, and liability.

JPMorgan Chase updated its terms in late October to prohibit third-party agents that access accounts using customer credentials. Bank of America also warned customers against sharing login information with AI agents, reflecting concerns over security and the potential consequences of automated account activity.

Liability Shifts Toward Customers

Other institutions are adopting similar approaches. Citi revised its online user agreement in November, making customers responsible for activity conducted by third-party AI agents they authorize, including situations involving compromised credentials or account hacking. The bank distinguishes these external tools from its own AI services, which operate under different terms.

PNC updated its terms on November 18, while Citizens reportedly communicated to customers that they could be responsible for actions taken by their AI assistants.

Wells Fargo’s approach also emphasizes the importance of formally revoking an agent’s access. Under the reported terms, customers may remain responsible for an agent’s activity unless they can demonstrate that access was revoked. This raises practical questions about how customers can establish that an AI system has stopped interacting with their accounts and what protections may apply if it continues operating after authorization is withdrawn.

The Accountability Challenge in Agent-Driven Finance

The emerging policies highlight a complex accountability issue. Customers may authorize an AI agent to accomplish a specific objective without approving every individual action the system takes to complete it.

Errors could involve duplicate purchases, unintended transactions, or spending outside the user’s intended parameters. In such circumstances, responsibility may be disputed among the consumer, the AI provider, the financial institution, and the merchant.

Existing rules governing electronic transfers, account access, and data sharing provide a regulatory foundation, but the treatment of transactions initiated by AI agents remains an evolving issue. Regulatory scrutiny is likely to increase as agents gain greater autonomy over financial activities.

Banks Build Their Own AI Ecosystems

The restrictions on external AI agents are occurring alongside significant investment in banks’ proprietary AI tools. Wells Fargo reports substantial usage of its generative AI assistant, while Bank of America and Citigroup have continued expanding their own AI-powered capabilities.

The strategic distinction is increasingly clear: banks are seeking to capture the efficiency and convenience benefits of AI while maintaining control over authentication, customer interactions, data security, and transaction execution.

This approach could create a more controlled environment for AI-driven banking services, although it may also limit interoperability between financial institutions and independent AI platforms.

Closing Insights

The growing restrictions on third-party AI agents represent an important development in the evolution of digital banking. As AI systems move from providing information to executing transactions, financial institutions are reassessing the balance between convenience, security, and liability.

Investors and industry participants should monitor regulatory guidance, consumer protection developments, and the evolution of bank-owned AI platforms. The long-term competitive opportunity may depend not only on the sophistication of AI agents but also on the institutions capable of establishing reliable authorization frameworks, transparent liability policies, and customer trust.

 

For a confidential discussion regarding retail banking strategy, insurance distribution models, customer loyalty ecosystems, digital financial services, or cross-border financial innovation opportunities, contact our senior advisory team.

Leave a Reply

Your email address will not be published. Required fields are marked *

More like this

Seraphinite AcceleratorOptimized by Seraphinite Accelerator
Turns on site high speed to be attractive for people and search engines.