AI agent for business: who uses agents, what they do, and what businesses spend

Organizations alongside $1 milliard or additional in income are scaling AI agents at nearly twice the charge of those below that line: 40% versus 22%. An AI delegate is application that completes multi-step tasks, specified as answering client questions or compiling research, alongside minimal individual input.

Customer assistance and investigation are the most average jobs for agents, and norm is the top obstacle for teams putting them into production. Hostinger’s own AI aide resolves 91% of concerning 1.5 myriad assistance conversations a duration without individual involvement.

Key AI delegate data for endeavor in 2026

Here are the key numbers on AI agents in business:

1. 40% of organizations alongside $1 milliard or additional in income are scaling AI agents in at smallest one function, compared alongside 22% of smaller organizations.
2. 62% of ample US organizations are building, deploying, or evolving AI agents, up from 53% a fourth earlier.
3. 46% of surveyed US small director firms use AI, and another 15% scheme to commencement inside a year.
4. 24.6% of all US director businesses use AI in a endeavor function.
5. 26.5% of delegate builders name client assistance as their chief use case, the most average answer, onward of investigation and data inspection at 24.4%.
6. 58% of leaders at ample US organizations study measurable endeavor value from AI.
7. 22% of total AI value comes from agents, up from 17% the twelvemonth before.
8. $29.2 billion is the forecast for worldwide spending on AI agents and assistants in 2026, up from $16.5 milliard in 2025.
9. 32% of delegate practitioners name norm as their top obstacle to production.
10. 87% of customers say companies using GenAI in client assistance must recommendation a way to attain a human.

Larger companies guide AI delegate acceptance in 2026

In 2026, 40% of respondents at organizations alongside $1 milliard or additional in income stated they were scaling AI agents in one or additional functions, compared alongside 22% at smaller organizations, according to a McKinsey survey. Smaller current method under $1 milliard in revenue, which includes many midsize companies. 

The large-company portion was 27% in 2025, during the smaller collection stayed flat. About two in 10 respondents stated their institution had reached the scaling phase throughout the entire organization, a stricter measure than scaling in one function.

KPMG’s AI Pulse survey established 62% of 314 US C-suite and endeavor leaders at organizations alongside $1 milliard or additional in income building, deploying, or evolving AI agents, up from 53% a fourth earlier. And 25% were evolving or implementing multi-agent systems, up from 6% in all of the two quarters before.

In a LangChain survey of 1,340 professionals, most of them in innovation roles and already operating alongside agents, 57.3% had agents in production, and another 30.4% were evolving them alongside tangible plans to deploy. That’s a lesser bar than McKinsey’s scaling phase, and the respondents lean toward first adopters, so the portion runs higher. 

Size made a difference here, too: 67% of organizations alongside 10,000 or additional workforce had agents in production, compared alongside 50% of those alongside small than 100 employees, although another 36% of the smaller collection were evolving them

Across the surveys, smaller organizations are operating on agents nearly as frequently as ample ones, but small have moved them into manufacturing or scaled them.

Customer assistance and investigation guide AI delegate use cases

Customer assistance and investigation are nearly tied as the most average jobs for AI agents: 26.5% of respondents in LangChain’s study named client assistance as their chief use case, and 24.4% named investigation and data analysis. Between them, the two document for additional than fractional of chief deployments, and another 18% use agents chiefly to automate inner workflows.

Big companies lean toward inner uses. At organizations alongside 10,000 or additional employees, inner efficiency tops the use-case catalog at 26.8%, onward of client assistance at 24.7% and investigation and data inspection at 22.2%.

McKinsey ranks delegate activity alternatively since it asks a distinct question: anywhere agents have reached the scaling phase, not all team’s chief use case. By that measure, respondents most frequently measure agents in data innovation (IT), cognition management, and application engineering, during respondents in person products and retailing most frequently use them in promotion and sales.

Coding agents are already shifting several purchasing decisions. About 20% of respondents stated their institution was scaling application coding agents, rising to 31% at larger enterprises. And 32% of respondents at organizations using AI stated they had skipped purchasing at smallest one application merchandise or characteristic since it could be built in-house alongside agentic coding tools, most frequently in innovation and healthcare. 

The answer depends on what you measure: client assistance and investigation are the most average chief jobs for agents, during IT and application engineering are anywhere they’ve scaled furthest.

Small businesses use AI mostly for penning and marketing

Of the US small director firms that use AI, 83% use it for penning or marketing, fine onward of idiosyncratic efficiency at 61% and preparedness or inspection at 51%, according to the Federal Reserve Banks report. 

Across all surveyed firms, 46% stated their endeavor or workforce use AI, another 15% planned to commencement inside 12 months, a third had no plans to use it, and 6% were unsure. These figures shield AI in general. None of the surveys current measure AI delegate use at small businesses specifically.

AI users study gains: 71% saw higher efficiency and 31% higher sales, during 77% saw no alter in labor expenses and lone 16% saw them fall. Most of these firms are motionless early, though. About fractional of AI users are experimenting, 44% have partially built AI into their endeavor processes, and 7% have completely unified it.

Firms preparedness to commencement citation applicable hurdles: 54% can’t discover tools that fit their needs, and 37% deficiency the period to execute AI or train staff. Starting alongside one workflow you already run narrows the hunt to a sole job, as our small endeavor AI guide shows.

US Census Bureau data shows lesser AI use than the Fed survey, and a apparent increase since spring. In the Business Trends and Outlook Survey, 24.6% of US director businesses used AI in a endeavor function in the duration ending October 18, 2026, up from 19.8% in the duration ending May 3.

The smallest firms are following the identical trend. Use at firms alongside one to four workforce rosy from 19.0% to 24.3% complete the identical months.

Larger firms use AI more. As of October 18, use reached 35.3% at firms alongside 100 to 249 workforce and 43.6% at firms alongside 250 or more.

The two surveys aren’t immediately comparable. The Fed asks whether a endeavor or its workforce use AI at all, according to a late-2025 example of small firms, during the Census Bureau asks whether a endeavor used AI in any endeavor function in the former two weeks, throughout director businesses of all sizes. 

Both item the identical way, though: AI use at small firms is growing, but most are motionless first in adopting it.

Agents document for 22% of the value companies get from AI in 2026, up from 17% in 2025, according to BCG’s Applied AI Index. The indicator draws on a study of additional than 1,300 elder executives throughout 20-plus sectors.

Maturity makes a big difference. 44% of the most AI-mature companies, which BCG calls future-built, already get value from agentic AI, compared alongside 2% of laggards.

McKinsey’s study shows a akin pattern. AI elevated performers, the approximately 6% of respondents who trait at smallest 5% of EBIT (earnings before involvement and taxes) to AI and study important value, were additional than three times as apt as others to be scaling agents in most endeavor functions. The study doesn’t say whether scaling agents drives the results or powerful performers merely measure more.

Measurable value from AI general is additional common. In KPMG’s survey, 58% of leaders at US organizations alongside $1 milliard or additional in income reported measurable endeavor value from AI initiatives, including but not constricted to agents. Productivity gains led at 55%, followed by faster decision-making at 49%, improved client and employee experiences at 38%, and stronger financial achievement at 37%.

Hostinger’s own assistance array shows what this looks akin in one operation. Hostinger’s AI aide resolved 91% of approximately 1.5 myriad assistance conversations a month without individual involvement. That assistance activity now runs inner Hostinger Agent, which combines assistance alongside specialized agents for endeavor tasks.

Spending on AI agents and assistants is forecast to attain $29.2 milliard in 2026

Worldwide spending on AI agents and assistants volition attain $29.2 milliard in 2026, up from $16.5 milliard in 2025, according to Gartner’s forecast, growth of concerning 77% calculated from its table.

Source: Gartner, September 2026.

Gartner expects spending to additional than twice again in 2027, to $65.5 billion. Part of that reflects what the forecast now counts: it includes person agents and assistants alongside endeavor spending.

Even so, agents remain a small piece of total AI spending. The 2026 fig is concerning 1.1% of total worldwide AI spending forecast for the year, calculated from the identical forecast. None of these sources puts a representative cost on a endeavor AI agent.

Cost holds rear a smaller group of organizations so far. In McKinsey’s survey, concerning 10% of respondents stated expenses had constricted their organization’s use of AI agents, the identical portion as for chatbots and coding agents. For AI overall, 20% stated functioning costs, including tokens, had held them back.

Overall AI budgets are climbing fast. BCG established company AI spending doubled in a twelvemonth to 3.3% of revenue, throughout all AI.

Large companies are adding disbursal controls. In KPMG’s survey, 74% of leaders contain disbursal reviews in their AI endorsement processes, up from 61% the former quarter, 70% use AI monitoring dashboards, and 43% set use or token budgets.

Quality is the biggest obstacle to putting AI agents into production

Quality is the biggest obstacle to putting agents into production: 32% of respondents in LangChain’s study named it first, onward of latency at 20%. Quality current method accuracy, relevance, consistency, and staying on spirit and inside policy. 

Cost came up small frequently than in former years, and at enterprises alongside 2,000 or additional employees, safety classified second at 24.9%.

Most teams detect what their agents do

In the identical survey, most teams track what their agents do:

• 89% of respondents have several form of observability for their agents.
• 52.4% run offline evaluations on test sets.
• 37.3% run online evaluations in production.

Large organizations are environment limits, but complete controls are rare

Large companies are fencing off decisions. In KPMG’s survey, 49% of leaders had defined high-risk use cases anywhere agents may not decide on their own, and 73% felt assured in their AI governance, up from 57% a fourth earlier. Few have the complete toolkit yet: BCG’s indicator established lone 5% of companies had all authority it lists as crucial for AI agents.

Security policies lag rearward use: 82% of surveyed organizations already use AI agents, but lone 44% have policies in location to safe them, according to Hostinger’s agentic AI statistics.

Customers anticipate a way to attain a human

Customers diagram the clearest line. In a Gartner survey, 87% stated companies using generative AI (GenAI) in client assistance must recommendation a way to attain a individual agent, equal although 50% stated GenAI makes their interactions easier. This covers GenAI in client service, not agents specifically. 

What forecasts say concerning AI agents through 2030

These are projections from consultancies and analysts, not measurements, and all uses its own definitions.

Gartner’s cancellation forecast cites escalating costs, unclear endeavor value, or inadequate hazard controls. Only concerning 10% of McKinsey respondents say disbursal limits their use of agents today, so this is a prediction concerning forthcoming project failures, not a current constraint.

How to commencement using AI agents in your business

1. Start alongside one customer-facing or investigation task. These are the jobs agents already grip most often, so tools and examples are uncomplicated to find. If you’d fairly not build an delegate from scratch, Hostinger’s agentic AI tools contain a ready-made delegate for support, content, and SEO tasks, as fine as personal agents you run yourself.

2. Keep a path to a person. Nearly nine in 10 customers anticipate a individual choice whenever a business uses GenAI for service.

3. Check norm before you add a second task. Review a example of the agent’s answers for accuracy and tone, since norm problems halt additional agents from reaching manufacturing than item else.

4. Cap expenses from day one. Set a use bounds before launch, the identical authority many ample organizations now build into their AI approvals.

Most businesses are motionless first alongside AI, so one project done fine is a sensible location to start.

All of the tutorial satisfied on this website is topic to Hostinger's rigorous editorial standards and values.