FIS and Anthropic Expand Partnership to Deploy Advanced AI Agents in Banking


RECEIPT (New York Stock Exchange: FIS), a global financial technology provider and anthropica company focused on artificial intelligence security and research, announced the expansion of their strategic partnership. This development focuses on integrating advanced agency AI into core banking functions and focuses on tackling it first. financial crimes.

The centerpiece of this expanded cooperation is the launch of a specialized unit on Financial Crimes. Artificial Intelligence Representative.

Powered by Anthropic’s Claude models, the agent is designed to significantly speed up the fight against money laundering (AML) processes.

Traditionally, these investigations can take hours or even days due to the need to manually collect evidence from different systems.

New artificial intelligence The tool automates this by compiling comprehensive evidence packages across a bank’s core platforms, evaluating transactions against established patterns of illicit activity and prioritizing only the most suspicious cases for human review.

This addresses a pressing industry issue.

According to estimates, approximately $2 trillion in illicit funds circulate in the global financial system each year, while US institutions alone fork out $35-40 billion each year. AML adaptation efforts.

Much of this expense involves labor-intensive tasks that yield high rates of false positives.

By compressing research timelines into minutes, the agent aims to reduce costs, reduce manual workloads, and compatibility Teams need to focus their expertise on real threats such as money laundering linked to drug trafficking, terrorist financing and other criminal enterprises.

FIS is positioned to deliver this solution by serving as the trusted infrastructure layer that processes a significant portion of the world’s transactions.

Customer data FIS provides full traceability, auditability and regulatory compliance by remaining securely in controlled environments.

Anthropic’s Applied AI team and forward-deployed engineers are working side by side with FIS experts to co-develop the agent, facilitating knowledge transfer that will enable FIS to independently build and scale additional purpose-built agents in the future.

Early adopters include Bank of Montreal (BMO) and Amalgamated Bank, which are already involved in development and testing. Wider availability for FIS clients is planned for the second half of the year 2026.

The partnership envisages a broader roadmap beyond financial crimes, covering areas such as credit decisions. customer participation, fraud prevention and deposit retention; all running on a unified, managed AI platform.

Stephanie FerrisFIS CEO and President highlighted the transformative potential: Banks around the world are looking for AI systems that actively perform tasks rather than merely providing suggestions.

he told RECEIPT As the foundational intermediary that protects data and manages intelligent agents, positioning organizations to evolve into “agent-first” operations.

Jonathan PelosiFinancial Services Manager at anthropichighlighted the complementary strengths of the alliance.

While FIS contributes unparalleled regulatory understanding, customer relationships and transaction data, claude Provides advanced reasoning, clear explanations and safe operation in highly regulated environments.

Each agent’s conclusion is tied directly to verifiable source information, preserving human oversight for final decisions.

This partnership extension underscores a broader shift towards responsible AI deployment at the enterprise level finance.

By combining FIS’ deep domain expertise anthropicThe collaboration with ‘s leading AI capabilities aims to set a new standard for secure, effective and scalable intelligence in banking. As financial institutions deal with increased regulatory scrutiny and sophisticated threats, such innovations to develop security While increasing operational efficiency.





Source link

Leave a Reply

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