Goldman Says AI Agents Face a Crucial Test: Making Money

AI agents have quickly become Silicon Valley’s latest obsession. Now comes the harder part of figuring out how they’ll make money.

The attention comes amid Meta’s recent launch of Muse, which has brought fresh attention to a new generation of AI software designed to shop online, book travel, and complete tasks on users’ behalf.

Goldman Sachs says the technology is undergoing a fundamental change.

“I think we are at the beginning of a paradigm shift. We are going from a conversational relationship with these agents to a much more action-oriented relationship,” Eric Sheridan, business unit leader of Goldman Sachs Research’s technology, media, and telecommunications group, said in a post published Friday on the bank’s website.

For companies building AI agents, the next step is figuring out how to turn those capabilities into a sustainable business, and Sheridan said the answer is likely to look familiar.

“The mass market for AI agents will be monetized with advertising and subscriptions, much like the way the web operates today,” Sheridan said.

He said AI agents would likely offer premium, ad-free subscriptions alongside free, ad-supported versions.

Sheridan said AI and machine learning could make digital advertising more efficient by helping automate how ads are created, placed, and measured.

That business model will also depend on consumers trusting AI agents with sensitive information such as calendars, passwords, and payment credentials. Sheridan said broader access would allow the software to carry out increasingly complex tasks on the users’ behalf.

From chatbots to commerce

That shift marks AI’s move from the infrastructure layer to the platform layer, according to Sheridan.

“We think the advent of consumer-facing, action-oriented agentic commerce marks the emergence of the platform layer,” he said.

The focus on AI agents has intensified in recent weeks as Meta’s AI agent, Muse, climbed to the top of Apple’s US App Store free app rankings after its launch.

The release adds to a growing wave of AI agents from OpenAI, Anthropic, and others that can browse the web and complete tasks on users’ behalf.

The trend has also played out outside the US.

Earlier this year, OpenClaw — an open-source AI agent framework created by Austrian developer Peter Steinberger — became a viral hit among developers and gained particular traction in China, where businesses and developers adopted the framework to build AI agents.