The future of work will become increasingly agentic. Trust will be the gate to AI agent adoption in the years ahead. Along the way, industries and business models will be disrupted. As we near the end of 2025, one burning question remains: how will agent adoption play out?
John Naisbitt, author of “Megatrends”, fervently believed that “the most reliable way to forecast the future is to try to understand the present.” His mindset was that the future is embedded in the present.
And that is how we think about making AI predictions. We start with high-quality data snapshots of today, then project where they’ll take the market over the next few years.
What the data reveals -- and the predictions for tomorrow
In August 2025, G2 conducted a carefully designed survey of a representative sample of B2B software buyers and influencers. We also conducted more than 40 interviews with technology leaders across major industries to gain a deeper understanding of the context. Finally, we analyzed data from G2 Reviews across relevant categories. All of this research culminated in our newly published G2 Insight Report, A Leap of Trust: AI Agents are Winning Hearts and Wallets.
Here are five bold predictions for 2026 and beyond, based on our G2 Research findings:
1. Enterprise mega-budgets will expand the AI agents market.
In 2026, more than 35% of enterprise companies will have budgets of $5 million or more for agents, encompassing software, services, and staffing. About 10% of enterprise companies will allocate $10 million or more for agents, especially as their workflows mature and they grow through increased efficiencies.
In our latest report, we documented enterprise spending on agents across industries, along with an 83% satisfaction rate with their performance to date. Interviews revealed a growing desire to expand use cases and invest even more in data and talent. This points to another blockbuster year for agents, likely pushing them past the 40% compound annual growth rate (CAGR) projected by research firms such as markets.us.
2. The AI orchestration market will explode.
We predict that the AI orchestration market will triple in size by 2027 to more than $30 billion.
All signs point to increased enterprise reliance on multi-agent systems to boost action accuracy and outcome quality. However, this will require greater investment in AI orchestration as agents cross boundaries and platforms. Agents will also need tight orchestration with robotic process automation (RPA) systems and data repositories to maximize efficiency.
Recently, G2 launched the AI Orchestration category, which most sources estimate as a $10–11 billion market. Expect this software category to grow rapidly as enterprises seek conductor-like solutions to bring together many moving parts. While some market research firms project a $30 billion market by 2030, developments strongly suggest it will reach that size three years ahead of schedule.
Orchestration emerged as a top-three priority among our technology interviewees — and for some, a growing concern. Our report also revealed that more than half of the companies’ agents were messaging other agents outside their platforms or systems. This points to rapid expansion in agent-to-agent (A2A) and Model Context Protocol (MCP) adoption, along with related service offerings.
3. A new category will emerge to create third-party guardrails for AI agents.
By the end of 2026, more than half of enterprises will use third-party services to create and oversee guardrails for AI agents rather than relying on vendors or platforms to minimize risks and errors. The 65% CAGR predicted for this category will not only come true; it may prove to be an underestimate.
Our report revealed that nearly two-thirds of companies were surprised by the extent of oversight required to manage agents, despite vendor claims. Additionally, only half of the buyers held agent vendors in high trust regarding reliability claims. Combine that with the widespread number of severe incidents, and it’s easy to see why demand for third-party guardrail services is set to leap.
A growing number of startups now offer agent assessments, policy-recommendation services, developer tools, and related solutions — all pointing to a new layer in the agentic space: risk management.
4. Agent builder platforms will expand their lead over in-house builds.
By 2027, agent builder platforms will widen their lead over in-house builds, growing from a 3:1 ratio to 5:1. This shift will result from in-house build programs delivering disappointing total cost of ownership (TCO), coupled with a high failure rate. Platforms such as Agentforce, ServiceNow, and Copilot are advancing in skill offerings, outcome quality, and time-to-value.
Our report found in-house builds ranked last in satisfaction, time-to-value, quality of actions, and ease of use. Notably, almost one in four in-house launches produced no meaningful outcomes in the first year. While it makes sense on paper to control your own data and destiny by building an internal agentic system, in practice, it’s challenging to staff and keep up with the industry’s rapid pace of change.
This is good news for incumbent SaaS giants, as more than two-thirds are expected to offer agent builder platform capabilities by 2027.
5. Agent evaluators will become designers, elevating their value in the workforce.
By 2028, one in four AI agent evaluators (human-in-the-loop) will move into higher-value (human-on-the-loop) roles. As mentioned earlier, companies told us that AI agents required more oversight than expected based on current large language models (LLMs) and data capabilities. This required employees to evaluate agent actions before or after the fact, changing the nature of their jobs. As models and data improve, oversight requirements will decrease, creating higher-value opportunities for employees to leverage agents across workflows.
Our report documented how human gates on actions drive wider deployment and scale cost savings through trust. While agents eliminate repetitive, low-value work, the act of evaluating them will soon become “the new boring.”
As agents generate a digital labor force, new roles such as agent boss, domain expert, agent trainer, and agent auditor will emerge — likely filled by those who demonstrated judgment and expertise in evaluations. Much like how bank tellers evolved from money counters to customer service agents to financial advisors after the introduction of ATMs, the rise of agents will elevate human work experiences.
Agents are here for our growth constraints, not our jobs
We’re excited about what the rise of agents means for businesses of all types and for the future of the SaaS industry. While some worry that agentic disruption will make existing products obsolete, it’s clear that it will also spark the creation of new ones to replace them.
by Washija Kazim
by Sidharth Yadav
by Brynne Ramella
