How OpenAI, Anthropic, and Google Are Competing to Own the Future of Banking

Agentic AI is no longer a proof-of-concept in banking—it’s running in production. Within weeks of each other, HSBC appointed its first Chief AI Officer, Starling launched the UK’s first agentic AI financial assistant, and Revolut rolled out its AIR platform to 13 million UK clients. The race to embed AI into financial operations has moved from exploration to execution.

What Separates the Three Leading Platforms

The major AI providers are competing on different strengths, and in regulated industries those differences translate directly into what gets automated, how safely, and with what level of oversight.

• Claude (Anthropic): Built for precision in high-stakes, document-heavy work. Most effective for contract review, regulatory compliance checks, and decisions requiring a clear audit trail. Anthropic is currently limiting access to its more powerful Claude Mythos Preview, reflecting a strategy of controlled deployment over broad access.
• ChatGPT (OpenAI): The broadest operator. It reasons across multiple steps, integrates diverse data sources, and handles complex workflows autonomously—from loan qualification to transaction flagging. OpenAI is granting several additional UK banks access to GPT-5.5-Cyber, prioritizing scale over restriction.
• Gemini (Google): Optimized for large-volume document processing, with a 1-million-token context window. For banks analyzing large volumes of financial records or research, this means faster answers from material that would previously take teams days to work through.

Where Agentic AI Is Already Changing Finance

• AML investigations: FIS and Anthropic announced a collaboration to compress anti-money laundering investigations from hours to minutes, automatically assembling evidence and surfacing the highest-risk cases for review
• Loan approvals: Moving toward same-day decisions as agents pull together relevant data without manual review
• Personal finance management: AI agents that autonomously monitor spending, flag unusual transactions, and move surplus cash—representing a fundamental shift in the bank-customer relationship

The Risks That Come with Speed

The same qualities that make agentic AI valuable in finance introduce new risks. Without clearly defined boundaries, speed and autonomy become liabilities. AI hallucinations in financial environments can have immediate customer consequences. Data privacy—what the agent can see, how it’s used, where it’s stored—needs to be explicitly defined and disclosed to customers before they opt in.

Read the full article on TechBullion