12 Examples Of AI Agents Across Industries (2026)

Aditi Pareek· Senior SEO Analyst at Appic Softwares
Updated:
AI Agents

An AI agent is software that takes a goal, works out the steps, uses tools such as search, databases or payment systems, and acts, usually with a person approving the risky steps. A chatbot replies to a question. An agent also does the work the question points to, such as issuing a refund, drafting a clinical note or opening a pull request.

The 12 examples below are deployments with public numbers or documented limits. Where a figure comes from the company itself and nobody has audited it, we say so. Figures are current as of November 2026.

AI Agent Market Size And Adoption

A market forecast estimates revenue. It does not tell you whether the agents work, which is what the examples below try to show.

12 Examples Of AI Agents In 2026

1. Customer support: Klarna

Klarna launched an OpenAI-powered support assistant in early 2024. In its first month it handled 2.3 million conversations, two-thirds of all customer service chats, which Klarna said equalled the work of 700 full-time agents. It also reported 25% fewer repeat inquiries and a resolution time under 2 minutes, down from 11 (Klarna, 27 February 2024).

In May 2025 CEO Sebastian Siemiatkowski told Bloomberg that Klarna was hiring human support staff again. He said AI was cheaper but produced "lower quality" service, and that customers should always be able to reach a person (Entrepreneur). An agent can take most of the volume and still need a human route for the rest.

2. Customer support at a software company: Salesforce

Salesforce runs its own Agentforce agent on help.salesforce.com. About six months after the October 2024 launch it had handled more than 500,000 conversations, resolved over 84% of questions without a person and handed 4% to a human support engineer. An editor's note says it passed 1 million conversations by July 2025 (Salesforce). These are Salesforce's own numbers. The company says it keeps the handoff rate where it is on purpose, and routes renewal questions to a person.

3. Online shopping: Amazon Rufus

Amazon said more than 300 million customers used its Rufus shopping assistant in 2025 and that Rufus drove nearly $12 billion in incremental annualized sales, up from the $10 billion pace it cited after the third quarter (Amazon's Q4 2025 results, as reported by PPC Land). Amazon counts purchases customers might not have made without Rufus. It has not published how it calculates that, so read the figure as Amazon's estimate.

4. Autonomous driving: Waymo

Waymo said on 26 March 2026 that it was giving 500,000 paid rides a week across 10 US cities, double the 250,000 it reached in April 2025 (InsideEVs). The driving software plans routes and reacts to traffic with no driver in the seat. It operates in a short list of cities, not everywhere.

5. Farming: John Deere See & Spray

Cameras on the boom scan the field as the sprayer moves, identify weeds and fire individual nozzles, so herbicide lands only where it is needed. Deere said customers used See & Spray on more than 5 million acres in 2025 and cut non-residual herbicide use by nearly 50% on average, saving nearly 31 million gallons of herbicide mix (Deere release via Farms.com, November 2025). That figure is Deere's. Our guide to AI in agriculture covers other uses on the farm.

6. Clinical notes: Kaiser Permanente

The Permanente Medical Group offered an ambient AI scribe, which listens to a visit and drafts the note, to 10,000 physicians at 21 Northern California locations. In the first 10 weeks, 3,442 physicians used it across 303,266 patient encounters, and most of them saved about an hour a day at the keyboard. The results come from an NEJM Catalyst commentary, not a controlled trial, and a small share of notes contained AI errors that the team was still working to reduce (AMA).

7. Banking: Bank of America's Erica

Bank of America said in August 2025 that Erica, its virtual assistant, had passed 3 billion client interactions (Bank of America). Erica launched in 2018, before language-model agents, so it is closer to a narrow assistant than an autonomous agent, and a good example of how far a narrow assistant can scale. For agents built on newer models, see our guides to AI agents in finance, AI agents in insurance and companies using AI in fintech.

8. Cybersecurity: Google Big Sleep

In July 2025 Google said Big Sleep, an agent built by Google DeepMind and Project Zero, found CVE-2025-6965, a memory-corruption flaw in SQLite that Google said only threat actors knew about and were likely to exploit. SQLite fixed it in version 3.50.2. Google called it the first time an AI agent had stopped an exploit before it was used, but it did not say who the attackers were, so that claim cannot be checked independently (SecurityWeek). See also AI in cybersecurity.

9. Manufacturing: Siemens industrial agents

In May 2025 Siemens announced industrial AI agents coordinated by an orchestrator, building on its copilots for design in NX, production planning, engineering in TIA Portal (code generated from plain-language requests), plant operations and maintenance. Siemens says the agents already deliver measurable results in its plants and customer implementations and sets a target of up to 50% higher productivity. The release gives no figures and names no customers, so treat the 50% as a goal, not a result (Siemens).

Harvey says 80% of Am Law 100 law firms use its product, a figure it gives without a method, in the announcement of a $550 million raise at a $15.5 billion valuation (Harvey, 9 September 2026). Thomson Reuters launched CoCounsel Legal in August 2025 with guided agentic workflows and deep research (Legal IT Insider). Harvey's numbers cover adoption and funding. We found no published accuracy results for either product.

11. Software development: GitHub Copilot coding agent

GitHub made its Copilot coding agent generally available on 25 September 2025 (GitHub changelog). You assign it an issue, it works in the background and opens a pull request. It cannot mark its own pull request ready for review, approve it or merge it, so a person always gates the change (GitHub docs).

12. Education: Khanmigo, a cautionary result

A two-year randomized study led by Philip Oreopoulos of the University of Toronto followed low-performing students in 18 Tennessee middle schools. By year two, students using Khan Academy made faster math gains than the comparison group, but the researchers concluded the gains did not appear to come from the Khanmigo AI tutor, because most students did not engage with it productively. Students used it on about a third of the days they worked in Khan Academy, and many chats were off topic or attempts to get the answer (Chalkbeat, 25 August 2026). Khan Academy has since redesigned the interface. For more, read AI agents in education.

What The Examples Have In Common

The agents with clear results do one narrow job and measure it: resolution time at Klarna, hours at the keyboard at Kaiser, gallons of herbicide at Deere. Most keep a person in the loop. GitHub's agent cannot merge its own work, Klarna brought human agents back, and Khanmigo shows what happens when the tool is available and the student does not use it.

Before you build one, pick a single task with a number you can track, decide who approves the risky steps, and plan to review the data after launch. That matters given Gartner's forecast that more than 40% of agentic projects will be canceled by the end of 2027.

If you want help scoping or building an agent, Appic Softwares offers AI agent development covering AI-powered automation, machine learning and intelligent system development. Contact us with the task you want to automate.

FAQs

1. How Do AI Agents Improve Customer Service Efficiency?

They take repetitive questions and finish routine tasks without a person. Klarna reported 2.3 million conversations handled in the first month and resolution time cut from 11 minutes to under 2, and Salesforce reported over 84% of help-site questions resolved without a human. Both numbers are the companies' own. Klarna later started hiring human support staff again after saying AI alone produced lower quality, so the practical setup is an AI agent that handles the routine volume and hands off the rest, often alongside sentiment analysis AI to flag unhappy customers.

2. What Industries Benefit The Most From AI Agents?

The measured results in this article come from customer support, retail, transport, farming, healthcare, banking, security, manufacturing, legal work, software development and education. The strongest evidence is in narrow, repeatable tasks such as support chats, clinical notes and spraying herbicide. Broad claims about whole industries have thinner evidence.

3. Can AI Agents Completely Replace Human Workers?

Not in the deployments above. Klarna brought human support staff back, GitHub's coding agent cannot merge its own pull requests, and Kaiser Permanente's scribe produced a small share of notes with errors. Agents take over repetitive, data-heavy steps, and people keep the judgment calls and the sign-off.

Frequently Asked Questions

1. How Do AI Agents Improve Customer Service Efficiency?

They take repetitive questions and finish routine tasks without a person. Klarna reported 2.3 million conversations handled in the first month and resolution time cut from 11 minutes to under 2, and Salesforce reported over 84% of help-site questions resolved without a human. Both numbers are the companies' own. Klarna later started hiring human support staff again after saying AI alone produced lower quality, so the practical setup is an AI agent that handles the routine volume and hands off the rest, often alongside sentiment analysis AI to flag unhappy customers.

2. What Industries Benefit The Most From AI Agents?

The measured results in this article come from customer support, retail, transport, farming, healthcare, banking, security, manufacturing, legal work, software development and education. The strongest evidence is in narrow, repeatable tasks such as support chats, clinical notes and spraying herbicide. Broad claims about whole industries have thinner evidence.

3. Can AI Agents Completely Replace Human Workers?

Not in the deployments above. Klarna brought human support staff back, GitHub's coding agent cannot merge its own pull requests, and Kaiser Permanente's scribe produced a small share of notes with errors. Agents take over repetitive, data-heavy steps, and people keep the judgment calls and the sign-off.

Related Articles

AI Agent Development Companies
2/10/2025

Top 8 AI Agent Development Companies In 2025

As businesses increasingly adopt artificial intelligence (AI) to automate processes and enhance customer experiences, the demand for AI agents continues to rise. AI agents streamline operations, improve decision-making, and provide personalized industry interactions. Whether in fintech, real estate, or education, an AI development company is crucial in building intelligent, scalable, and industry-specific AI solutions. In […]

Read More
Agentic AI Frameworks
2/12/2025

8 Best Agentic AI Frameworks In 2025

With their ability to automate processes, connect to organizational systems, and perform specific tasks, agentic AI frameworks are becoming increasingly popular. Various AI agent features, such as automation, decision-making, and adaptive learning, make them highly efficient for businesses. The agentic AI frameworks provide an organized platform for deploying and overseeing these AI agents within a […]

Read More
Sentiment Analysis AI | Appic Softwares
2/7/2025

What Is Sentiment Analysis AI? 8 Ways To Use AI For Sentiment Analysis

Sentiment analysis AI is the process of identifying, extracting, quantifying, and analyzing affective states and subjective data from text using artificial intelligence technology, particularly natural language processing (NLP). Sentiment analysis’s main objective is to ascertain the attitude, feelings, or ideas that are conveyed in a certain text. These can be neutral, positive, negative, or even […]

Read More
Create An App Like Wanderlog
2/13/2025

How Much Does It Cost To Create An App Like Wanderlog?

In today’s fast-paced world, travelers seek seamless and efficient ways to plan their trips, making travel planner apps like Wanderlog increasingly popular. These apps help users organize itineraries, discover new destinations, find accommodations, and even manage budgets, all in one place. Whether it’s a solo adventure, a family vacation, or a business trip, a well-designed […]

Read More
AI App Development: How to Build AI and AI Agent Apps (2026 Guide)
6/6/2026

AI App Development: How to Build AI and AI Agent Apps (2026 Guide)

Discover how AI app development services help businesses build intelligent, scalable applications. Learn about development costs, technologies, use cases, and implementation strategies. Whether you're creating an AI chatbot, AI agent, or enterprise solution, this guide explains everything you need to know before starting your custom AI app development project.
Read More
AI Automation Companies
5/25/2026

Top 5 AI Automation Companies for Businesses In 2026

Discover the best AI automation companies helping businesses build AI-powered applications, automate workflows, improve customer engagement, and scale enterprise operations
Read More