Media Briefing: AI Is Changing Agency Economics Faster Than Contracts Can Keep Up

Digiday examined how generative and agentic AI are changing agency operations, pricing and client agreements, with Go Fish Digital President David Dweck providing perspective on how those changes are showing up in practice. Agencies are automating more marketing work and building proprietary AI technology, but the four agency executives Digiday interviewed aren’t overhauling their Master Service Agreements yet. Instead, they’re making targeted changes as specific needs emerge.

At Go Fish Digital, that has meant adding specific MSA clauses around brand safety and data protection over the past six months. Dweck also noted that Go Fish does not typically charge clients separately for its proprietary technology.

Key Takeaways

AI is already changing agency contracts, but mostly through incremental updates. Agencies interviewed by Digiday are using clauses, addendums and other targeted changes rather than rewriting their MSAs around AI. Changes can be triggered by client requests, use of proprietary agency technology or the introduction of an AI-powered tool for a specific project.

Go Fish has added protections around brand safety and data use. Dweck told Digiday that Go Fish has introduced specific MSA clauses addressing brand safety and data protection during the past six months, two areas receiving more scrutiny as AI becomes part of agency workflows.

Clients want clearer rules around what happens inside AI systems. The executives Digiday interviewed pointed to concerns including data storage and usage, human oversight, indemnity, disclosure of AI tools and ownership of intellectual property. Attorney Keri Bruce summarized one recurring client expectation: disclosure about which tools are being used.

AI efficiency is putting pressure on the traditional agency pricing model. Digiday notes that when AI reduces the hours required to complete work, it complicates a business model historically tied to hourly rates. Some agencies are introducing proprietary AI products, raising questions about whether and how technology should become a separate source of revenue.

Go Fish is keeping its proprietary technology within a familiar agency relationship. Dweck told Digiday that Go Fish doesn’t typically charge for its proprietary technology and would rather retain a familiar agency payment model than try to operate like a SaaS company.

The Go Fish Perspective

The contract discussion points to something broader than legal language. AI can change how work gets done well before a client-agency relationship is ready for a different commercial model.

That makes governance important at the operational level. As agencies introduce AI into more workflows, clients need clarity about where their data goes, which tools touch their work, how brand safety is protected and where human oversight remains. Go Fish’s recent MSA additions around brand safety and data protection reflect those immediate questions without assuming the economics or technology have settled into a permanent model.

There’s also a useful distinction between using proprietary technology to improve agency work and selling that technology as a separate product. Digiday’s reporting shows agencies are still working through that question. Go Fish currently sits on the former side: building and using proprietary technology without typically charging clients separately for it.

Read the Full Story

Digiday’s Kimeko McCoy looks at how AI is affecting agency contracts, cost structures and pricing models, with perspectives from Go Fish Digital, Mile Marker, Crispin, VML and Reed Smith.

Originally Published by Digiday: AI reshapes agency economics, but their contracts are still scrambling to catch up, September 10, 2026.