- The Shift Brief
- Posts
- Managing AI Costs: What Mid-Sized Firms Need to Know
Managing AI Costs: What Mid-Sized Firms Need to Know
The Shift Brief | Week of Dec 1st
Ryan Erickson
December 02, 2025
Most recent headlines about AI and finance focus on big institutions. Large banks announced new model partnerships. Global consultancies increased their investment in enterprise AI programs. Vendors continued to position their platforms as the future of financial technology.
But the most useful signal came from a quieter source: a new Forbes Finance Council article about how smaller and mid-sized financial institutions can manage rising AI costs.
This is the conversation that actually reflects what most firms are experiencing. Budgets are tight. Talent is stretched thin. Expectations around AI are increasing faster than internal capacity. The question is not whether AI matters. The question is how to adopt it without creating new cost burdens or operational risks.
The Forbes Takeaway: A Practical Path for Mid-Sized Firms
The Forbes piece highlights a set of challenges and strategies that are becoming common across the mid-market:
AI costs are rising quickly. Even simple use cases can require compute, data preparation, compliance involvement, and ongoing monitoring. Many firms underestimate the total cost of ownership. It's not just the cost of the tool. It's the cost of everything that surrounds it.
Smaller institutions don't need to replicate what global banks are doing. The article encourages firms to avoid building internal AI infrastructure unless they have the scale to support it. Self-hosting models, managing cloud compute, and maintaining model governance are expensive. A hybrid approach is often more sustainable.
Governance and compliance must stay central. For many firms, regulatory exposure is the biggest blocker. AI that creates uncertainty around data lineage, record keeping, or supervisory responsibilities introduces more risk than reward. Any adoption strategy needs governance built in at the start.
Focus on targeted, high-impact workflows. Forbes calls out that the best returns come from applying AI to the daily tasks that analysts, portfolio managers, and operations teams already perform. Examples include document review, note summarization, data transformation, and internal communication.
Vendor selection is more important than feature selection. The piece stresses that firms should evaluate vendors on transparency, reliability, and long-term stability. The goal is not the flashiest model. The goal is a safe and predictable path to value.
Overall, the message from Forbes is clear. Mid-sized financial institutions want AI, but they want it in a form that aligns with the realities of their team sizes, budgets, and regulatory expectations.
These stories show a widening gap. Large institutions are scaling AI aggressively. Smaller firms are trying to adopt it responsibly. The next year will likely determine which strategies prove most durable.
~Ryan Erickson, Founding Executive
TLDR: What you need to know about AI in Finance
HSBC partners with Mistral AI. HSBC announced a multi-year collaboration with Mistral to integrate generative AI across financial analysis, risk assessment, and client communication. The bank plans to self-host models inside its own infrastructure.
Accenture expands its deployment of ChatGPT Enterprise. Accenture is equipping tens of thousands of employees with ChatGPT Enterprise as part of its broader AI adoption strategy. The company expects significant operational efficiency gains as it integrates AI into consulting workflows.