84% of enterprises plan to boost AI agent investments in 2026

The best thing about the AI revolution isn't the productivity gains, time savings, or even unlocked opportunities. It's that fans of The Matrix get to finally live in a world with artificially intelligent computer programs called agents.

While AI agents don't wear suits (yet), they're the hottest technology in the SaaS world right now. But have businesses actually bought in? 

To explore the current state of agentic adoption, Zapier surveyed over 500 enterprise leaders to find out how their companies are developing, deploying, and investing in AI agents. 

Key findings

• Close to 3 in 4 enterprises (72%) are now using or testing AI agents.

• 49% of customer support and 47% of operations teams have deployed AI agents.

• Nearly half of enterprises (47%) use AI agents for data management.

• Human-in-the-loop is the most popular approach to AI agents in enterprises.

• Over 4 in 5 enterprise leaders (84%) say it's likely or certain they'll increase AI agent investments over the next 12 months.

• 30% of leaders say they see the most potential for AI agents in automating routine workflows. 

Table of contents:

• 3 in 4 enterprises are now actively using or testing AI agents

• Customer support and ops teams are most likely to use AI agents

• Enterprises are taking a human-in-the-loop approach to AI agent management

• 84% of enterprises will likely increase AI agent investments in the coming year

• Connect AI agents to your work with Zapier

3 in 4 enterprises are now actively using or testing AI agents

Agentic AI is more (a lot more) than just the tech world's favorite buzzword. It's become a full-fledged business reality, and 72% of enterprises now have AI agents deployed and operating autonomously. 

Of this group, 40% have multiple AI agents in production, and 32% are still in the pilot or testing phase. Another 14% say they plan to adopt AI agents even if they don't have any right now. This makes for a total of 86% of enterprises with AI agents in production, pilot, or planning phases.

Adoption is widespread, but businesses are split on how they're approaching agent development. Just over half of enterprises (53%) use major cloud provider platforms like AWS, Azure, and GCP to build AI agents—while just under half are leaning on open-source tools (48%), enterprise AI platforms (48%), or orchestration frameworks (46%).

None of these approaches is exclusive, meaning many enterprises are creating and deploying agents across multiple platforms or services. On the other hand, a much smaller group (26%) are coding custom agents from scratch. 

Customer support and operations teams are most likely to use AI agents 

Across the enterprise, Customer Support is the department most likely to have deployed AI agents (49%). Operations comes in at a very close second (47%), followed by Engineering (35%) and Marketing (31%). 

Sales and Finance are the teams least likely to use AI agents, but still come in at 26% and 24%, respectively.

While customer service might make you think of chatbots, AI agents can do a lot more than field complaints or connect you with a human representative. The most popular AI agent use case across departments is data management: 47% of enterprises now use agents for tasks like data entry and extraction.

Other top uses include document analysis and summarization (41%), customer support triage and response (41%), and report generation (36%).

Enterprises are taking a human-in-the-loop approach to AI agent management

Let's be clear about this: AI agents may be able to act autonomously, but they're not the ones running the show. Most businesses still prefer to keep a fairly tight leash on their bots, and heading into 2026, human-in-the-loop remains the most common approach (38%) to AI agent management. 

Human-in-the loop means building approval gates directly into AI workflows. So, when an AI agent drafts an email or spins up a data report, a real human being has to step in and review before it moves down the pipeline. With Zapier, you can automate this process, directing Agents to wait for human input and flagging team members automatically when it's time to enter the loop. 

But while there's no replacing human judgment, AI agents are getting smarter, and 1 in 5 (20%) enterprise leaders say their AI systems now operate autonomously with minimal oversight. 

At the same time, it's unlikely businesses will fully hand the reins over to AI  anytime soon, as these same leaders say security and data privacy concerns (18%) are the biggest barriers delaying AI agent adoption. 

84% of enterprises will likely increase AI agent investments in the coming year

You may not be bullish on agentic AI, but your industry probably is. Over 4 in 5 enterprise leaders (84%) say it's likely or certain their organization will increase AI agent investments over the next 12 months (48% likely, 36% certain).

On the other end of the spectrum, 2% say they're certain they won't increase investments in AI agents, and only 1% say they're actively decreasing funding next year. In other words, expect to see more AI agents deployed across more departments in 2026.

AI agent technology is a growing field, and most enterprise leaders are hopeful that it'll bring real change to their businesses. Close to 1 in 3 (30%) say they see the most potential for AI agents in automating routine workflows. 

Others see the highest potential for agents to improve customer experiences (17%) or assist with strategy and decision making (12%). Just 5% of enterprise leaders say they don't see potential for AI agents.

Bring AI agents to the enterprise with Zapier

If all this talk of throwing money at agents has you worried about tool sprawl and rising AI costs, Zapier lets you design and deploy AI agents within the context of the tech stack you already use. 

Instead of paying for new apps, external developers, or engineering hires, anyone across an enterprise can build no-code agents in minutes using Zapier templates and natural language prompts, then add them seamlessly to automated flows.

Zapier keeps agents out of silos with integrations to more than 8,000 apps and AI tools, so every department can integrate AI agents into their existing workflows.

Related reading:

• 78% of enterprises are struggling to integrate AI with legacy systems

• AI in business: Statistics and use cases

• Only 50% of enterprise leaders say AI benefits are widespread

• 34 enterprise AI statistics