Software Market Forecast to Expand 20% as AI Agents Boost Productivity

The next milestone in the development of generative artificial intelligence (AI) will be the unlocking of business productivity through AI agents that can perform sequences of actions with minimal intervention by the user, according to Goldman Sachs Research. This shift will expand the software market over the next several years. 

The market for customer service software—incorporating traditional software-as-a-service (SaaS) products and new AI agents—could expand by an additional 20% to 45% by 2030, writes Gabriela Borges, who covers emerging software in Goldman Sachs Research, in the team’s report. That’s compared to a scenario without a generative AI boost. Their analysis is based on value- and cost-based pricing methodologies, and their estimates are based on conversations with industry experts and pricing specialists.

The total addressable market for the broader software industry is expected to expand at least 20%, the researchers find, using the growth in customer service software as a “low-end proxy” for the sector. In fact, there may be greater potential for the market to expand in areas that are more directly tied to revenue generation, such as sales and marketing, as compared with customer service, which is primarily viewed as a cost center. Growth in the market for developer tools may benefit from the faster pace of innovation, and there may be idiosyncratic opportunities in areas such as security operations.

“We believe agents will drive productivity, and software companies will capture a portion of this value,” Borges writes. Goldman Sachs Research estimates the application software market could grow to $780 billion by 2030, a 13% compound annual growth rate from this year.

How AI can boost productivity

The researchers examined how AI agents are beginning to boost productivity. “While the majority of examples that we discovered in our industry diligence over the last six months could be described as chatbots with basic integrations to LLMs, we did find select examples of more advanced AI that support much more interesting use cases,” Borges writes. In many instances, these are either proof-of-concept activities or agents trained for internal use at software companies, but there is clear potential for commercialization longer-term. Large language models, or LLMs, are a type of AI that are trained on large amounts of data and can process and generate human-language content.

“Our technical deep dive illustrates the potential for agents to become the new user interface for knowledge workers,” Borges writes. By 2030, the agent portion of the software market may account for more than 60% of the total. In other words, the profit pool is going to shift to agents, but the entire market for software will be larger.

How are software companies using AI?

To date, even the definition of “agent” in the context of AI innovation has been somewhat difficult to pin down, and Goldman Sachs Research examines how different companies use the term. The industry is coalescing on a definition that centers on autonomy: Agents need to be non-deterministic, respond and be proactive to changes in their environment, and be able to remember context to effectively process workflow, Borges writes. 

Most deployed enterprise AI systems to date use the new technology as a type of pattern-matching tool, generating responses based on large data sets, or as a decision-assistance tool or software wizard. Agentic AI solutions that can execute tasks autonomously would represent a new level of innovation with significant potential to improve productivity.

Seen another way, the agent is where the LLMs combine with workflows and application programming interfaces (APIs) to perform tasks autonomously. 

Generative AI technology is improving, the researchers explain, and software companies are working to build the ecosystem needed to address what have been limiting factors so far, including the lack of a stable AI-enabled platform to build on. History suggests that application adoption naturally follows after standardization of the platform layer, Borges says, and such standardization is at least 12 months away.

Developers and customers still have concerns that range from data integrity to security and authentication, but these will become less of an obstacle as the platform layer matures. Other limitations holding back adoption are also being addressed, with work to improve reliability and memory for AI tools, for example.

This material is provided in conjunction with the associated video/audio content for convenience.  The content of this material may differ from the associated video/audio and Goldman Sachs is not responsible for any errors in the material. The views expressed in this material are not necessarily those of Goldman Sachs or its affiliates. This material should not be copied, distributed, published, or reproduced, in whole or in part, or disclosed by any recipient to any other person. The information contained in this material does not constitute a recommendation from any Goldman Sachs entity to the recipient, and Goldman Sachs is not providing any financial, economic, legal, investment, accounting, or tax advice through this program or to its recipient. Neither Goldman Sachs nor any of its affiliates makes any representation or warranty, express or implied, as to the accuracy or completeness of the statements or any information contained in this material and any liability therefore (including in respect of direct, indirect, or consequential loss or damage) is expressly disclaimed.

© 2025 Goldman Sachs. All rights reserved.