AI and Open Source in Financial Services: Mitigating Machine-Scale Cyber Risks, Governing Agentic Workflows, and Co-Engineering the Next Wave of Industry Standards

Abstract

What is the true cost and security risk of deploying AI in banking? The financial services sector faces unprecedented management challenges as AI integrates in core production. Authored by FINOS and LF Research, this report synthesizes critical insights from industry executives, regulators, and security experts in attendance at the AI in Finance Leadership Summit that took place in London in June 2026. Key issues include: An 18-to-24-month patching window: Rapidly evolving machine-scale security threats require firms to automate vulnerability remediation before legacy flaws are actively exploited. The rise of Tokenomics: Autonomous, agentic workflows run continuous loops that can quickly drain compute allocations, turning AI cost governance into an urgent boardroom priority. Pre-competitive collaboration: Because financial institutions share the same foundational software and regulatory oversight, creating proprietary compliance frameworks offers zero competitive advantage. Collaborative open source standards are required to protect enterprise margins. The report concludes with some frameworks to optimize token spend, secure application supply chains, and build resilient infrastructure.