Citigroup CEO Jane Fraser calls for strict controls on agentic AI
- Citigroup CEO Jane Fraser mandates strict controls for agentic AI
- Bank implements constant tracking of AI agents
- Authentication protocols emphasized for banking security

*this image is generated using AI for illustrative purposes only.
Citigroup CEO Jane Fraser stated that agentic artificial intelligence in banking must be governed by rigorous controls and authentication protocols. The bank is implementing constant tracking of these autonomous agents to mitigate operational risks.
Fraser highlighted the necessity of maintaining strict oversight as financial institutions integrate more advanced AI capabilities. The focus remains on ensuring that authentication mechanisms are robust enough to handle the complexities introduced by agentic systems.
Focus on security and compliance
The implementation of agentic AI requires a shift from traditional static models to dynamic, autonomous systems. Citigroup’s approach involves continuous monitoring to detect anomalies and ensure that each agent operates within predefined safety boundaries. This strategy addresses the growing concern regarding the integrity of automated decision-making processes in high-stakes financial environments.
How might regulatory bodies like the SEC or OCC adapt their compliance frameworks to specifically address the liability of autonomous AI agents in banking?
What specific technical standards for 'continuous monitoring' are emerging as industry benchmarks to prevent unauthorized actions by agentic AI systems?
Could the rigorous control protocols required for agentic AI create a competitive barrier that favors large banks over fintech startups with less robust infrastructure?

































