Anthropic, a major developer of advanced artificial intelligence models, is teaming up with financial software provider Fidelity National Information Services Inc (NYSE:FIS) to develop automated tools aimed at policing the international banking system. According to people familiar with the arrangement cited in reporting, the collaboration focuses on creating AI agents that can function with limited human supervision.
Under the partnership, the specialized agents will combine FIS’s large financial data sets with Claude, Anthropic’s sophisticated large language model. The stated aim is to build a system capable of investigating individuals and networks accused of exploiting financial channels - including drug traffickers, terrorists, and other criminal actors - by pulling together disparate records and data streams.
Market reaction was immediate: FIS stock rose 6.6% in after-hours trading on the Monday the report surfaced. FIS’s CEO Stephanie Ferris described the planned financial-crimes bot as a system that will independently gather evidence from varied data repositories and account histories. While Ferris said the AI should materially reduce the time and cost of conducting individual investigations, she was explicit that human investigators will continue to render all final case decisions.
Early adoption is slated to include the Bank of Montreal (NYSE:BMO) and Amalgamated Bank (NASDAQ:AMAL), which are expected to be among the first institutions to deploy the agent. The companies anticipate broader availability in the second half of the year, after embedded Anthropic engineers complete further development work.
The effort comes as banks devote billions of dollars annually to anti-money-laundering programs to meet demanding federal requirements and regulatory scrutiny. At the same time, the rapid progress of large AI models has weighed on the market values of established software vendors. FIS’s stock has fallen by more than 25% so far this year, reflecting investor concerns that some traditional products could be displaced if large firms build proprietary AI tools.
However, the Anthropic-FIS collaboration indicates an alternative path: AI research labs may elect to integrate their models with existing enterprise software, accelerating industry-specific capabilities rather than rendering them obsolete. The partnership aims to leverage FIS’s dataset strengths together with Anthropic’s model to speed investigative workflows while preserving human oversight in decision-making.
Summary
Anthropic and FIS are working to combine an advanced language model with extensive financial data to create semi-autonomous agents for financial-crime investigations. The tool is intended to collect and correlate evidence across multiple data sources, reduce the time and cost of investigations, and maintain human control over final outcomes. Initial pilots will include the Bank of Montreal and Amalgamated Bank, with general availability expected in the second half of the year.
Key points
- Partnership pairs Anthropic’s Claude model with FIS’s financial datasets to build AI agents for investigating criminals exploiting financial networks.
- FIS shares rose 6.6% in after-hours trading after the report; the company’s stock has fallen over 25% year-to-date.
- Initial rollouts will include Bank of Montreal and Amalgamated Bank; wide availability is expected in the second half of the year following engineering work.
Risks and uncertainties
- Timing risk - broad availability is projected for the second half of the year, a timeline that is conditional on completion of development work by embedded Anthropic engineers.
- Market risk for legacy software providers - investors worry that established products could be displaced as advanced AI models proliferate, a factor reflected in FIS’s year-to-date share decline of over 25%.
- Operational and compliance limits - although the AI is expected to lower time and cost per investigation, human investigators will retain final decision authority, meaning the technology may not eliminate substantial compliance spending by banks.