Observed Signal · Nov 18, 2026 · Policy Update · Source: techcrunch · Impact: 2/5 · Sentiment: Negative

Citi Updates Terms to Exclude Liability for AI Agents

Executive Signal Summary

Citibank, the third-largest US commercial bank, changed its terms of service this week, stating that customers using third-party AI tools for personal finance are responsible for any consequences, including losses from hacking. This follows similar policy updates by JPMorgan and Bank of America last month, and PNC changes its terms the same day. The bank also warned that if an AI tool leaks credentials, customers are liable for unauthorized transactions. This trend reflects financial institutions' concerns about unreliable AI agents accessing accounts, and it follows the recent launch of Anthropic's read-only personal finance tool for its AI assistant. Banks see AI agents both as a security risk and a competitive threat, aiming to control this emerging interaction pathway.

Polaris7 AgentPolaris7 Strategic Assessment
High Confidence

This article documents a significant shift in how major financial institutions are responding to the rise of AI agents, signaling potential friction for AI adoption in personal finance and highlighting liability and trust concerns that could impact the broader adoption of AI agents in commercial contexts.

MARKET SIGNALS

Track JPMorgan Chase Signals & Strategic Shifts

Polaris7 connects regulatory filings, primary sources, leadership changes, and M&A activity into structured intelligence. Set up your Polaris7 workspace to monitor relevant companies in your executive briefing.

  • Verified primary sources
  • Automated briefings
  • Instant watchlist setup

Key Takeaways & Evidence Grounding

  • Citibank updated its terms of service to hold customers liable for any actions by third-party AI agents accessing their accounts.
  • Citi also states customers are responsible if an AI tool leaks credentials, even if the account is subsequently hacked.
  • JPMorgan and Bank of America announced similar policies last month.
  • PNC changed its terms of service on the same day as Citi's update.
  • Anthropic's recent personal finance tool is explicitly 'read-only' and cannot move or invest money.

Connected Companies & Entities

“JPMorgan and Bank of America, the two largest banks in the country, announced similar policies last month.”

“JPMorgan and Bank of America, the two largest banks in the country, announced similar policies last month.”

“Anthropic's approach is telling: The personal finance tool it launched earlier this month is explicitly billed as 'read-only' and 'can't mov...”

Ontology Mapping & Concepts

Primary Source Grounding & Direct Attribution
Direct Origin Attribution
Primary Reporting: techcrunch•Published: Nov 18, 2026
Original Coverage Title: “More banks say customers using AI agents are on their own — even if they get hacked”

Related Market Signals & Shifts

Recent verified developments and strategic activity across this market segment.

AI AgentsNov 11, 2026

Banks Block AI Agents to Protect $1.6 Trillion Deposits

North American banks are blocking AI agents from customer accounts to protect low-cost deposits and margins. JPMorgan led the move on October 13, followed by Citi, Bank of America, and others. Wells Fargo will hold users liable for agent actions from November 24. The article highlights the financial stakes: $1.6 trillion in non-interest-bearing deposits, worth $47 billion annually if rates rise. Meanwhile, Stripe's Treasury for agents, powered by Fifth Third Bank, illustrates how banks can become wholesale infrastructure. The piece explores the 'agentic bank run' risk and outlines future scenarios for liability, margin capture, and banking's evolution.

Read assessment
AI Agents in BankingOct 18, 2026

US banks block AI agents from moving money

Major US banks and brokers are blocking AI agents that attempt to move money on behalf of customers, citing security, liability, and controllability concerns. Charles Schwab CEO Rick Wurster stated the firm blocks unauthorized agents weekly. JPMorgan Chase CEO Jamie Dimon warned against 'rogue agents,' while Citigroup and Bank of America restrict AI-agent-initiated transactions. Banks have contacted the Federal Reserve. Meta's agent Muse, launched in September, uses Plaid for financial data and is a key target. Analysts suggest banks also fear AI agents shifting funds to higher-yield accounts, costing billions in lost deposits. Meanwhile, banks are developing read-only MCP interfaces and their own AI assistants. Congress is debating liability for AI agent errors.

Read assessment
AI AgentsOct 23, 2026

AI Agents Bank Disruption Fears Overblown, Analysts Say

Wall Street analysts are pushing back against fears that AI agents could disrupt bank deposits by enabling consumers to easily move funds to higher-yield accounts. While Meta's Muse agent has sparked concerns, leading analysts like Gerard Cassidy (RBC) and Mike Mayo (Wells Fargo) argue that trust and convenience keep customers loyal. A Deloitte survey shows 72% of customers hesitate to share financial info with genAI tools. Banks like JPMorgan, Bank of America, and Schwab are taking protective measures and developing their own AI tools. No evidence of deposit shifts appeared in Q3 earnings, but risks remain for savings accounts.

Read assessment

WATCHLIST & BRIEFINGS

Receive future signals for JPMorgan Chase & competitors in your briefing

Stay ahead of material M&A deals, leadership changes, account wins, and product rollouts. Polaris7 connects primary filings, company newsrooms, and trade reporting directly into your executive briefing.

  • Regular primary-source checks
  • Structured Watchlist briefings
  • Preserves current context