Agentic AI, a new type of artificial intelligence with autonomy, is revolutionizing various industries. It can not only execute preset tasks but also think and make decisions proactively like a human, significantly enhancing work efficiency and decision-making quality.
Experts say that artificial intelligence agents【(Artificial Intelligence Agent), also known as agents, are software or hardware entities that can perceive the environment, make decisions, and take actions to achieve specific goals.They can autonomously respond to external stimuli, like an individual with intelligent behavior.】 will play a crucial role in software programming and cybersecurity, but they will also transform business workflows and business intelligence.
Agentic AI【refers to artificial intelligence with agency. This type of AI can not only execute tasks according to pre-programmed instructions but also set goals, plan, and take actions to achieve those goals like an autonomous intelligent agent. It has proactivity and can independently make decisions in complex and dynamic environments, rather than simply reacting passively to inputs.】is in the development stage, as proponents see the benefits of using autonomous AI agents to automate manual tasks across the organization.
In June, Agentic AI was rated by Forrester【Forrester Research is a globally recognized market research company focused on providing insights into the impact of technology on business and consumer behavior. Founded in 1983, it is headquartered in Cambridge, Massachusetts, USA. The company helps businesses make strategic decisions through research on market trends, consumer behavior, technological innovations, and more.】 as one of the hottest emerging technologies for 2025, taking generative AI further by emphasizing operational decisions rather than content generation. This approach is expected to impact business workflows, with organizations like Aflac【Aflac, founded in 1955 and headquartered in Columbus, Georgia, is the largest supplemental insurance provider in the U.S. and a component of the S&P 500 index】, Atlantic Health System【Atlantic Health System is a healthcare system serving New Jersey and New York, with several award-winning hospitals, including Morristown Medical Center, ranked #1 in New Jersey by U.S. News & World Report and Newsweek】, Legendary Entertainment【Legendary Entertainment is a leading media company with divisions in film, television, digital, and comics, dedicated to owning, producing, and delivering content to global audiences】, and NASA’s Jet Propulsion Laboratory【NASA’s Jet Propulsion Laboratory is a federally funded research and development center under NASA, located in Pasadena, California. Founded in 1936 by researchers from the California Institute of Technology, it is currently owned and sponsored by NASA and managed by Caltech.】 have begun exploring this technology.
CRM【CRM stands for Customer Relationship Management. It is a customer-centric business strategy and a set of corresponding technological means used to manage interactions between businesses and existing or potential customers, aimed at improving customer satisfaction, loyalty, and ultimately enhancing the profitability of the business.】 Leaders Salesforce【Salesforce is a leading global provider of cloud computing and customer relationship management (CRM) solutions. It provides features such as lead management, sales forecasting, opportunity management, and quote management, helping sales teams track potential customers more effectively, allocate resources efficiently, and improve sales performance.】 later shifted its strategic focus toward agent AI and announced the launch of Agentforce【Agentforce is a groundbreaking suite of autonomous AI agents launched by Salesforce to enhance employee capabilities and manage tasks in service, sales, marketing, and business, delivering unprecedented efficiency and customer satisfaction. Agentforce can help businesses scale their workforce on demand with just a few clicks. These AI agents can analyze data, make decisions, and perform tasks such as answering customer service inquiries, filtering sales leads, and optimizing marketing campaigns.】 IT service management giant ServiceNow【ServiceNow is a powerful enterprise software platform that occupies an important position in the field of IT service management. Initially started in a home office with a laptop, one employee, and a few volunteers, it has grown over the years to become a globally recognized provider of enterprise software services】 has also added AI agents to its Now Platform【Now Platform is a powerful enterprise cloud platform launched by ServiceNow, covering multiple business processes in IT service management (ITSM), human resources management (HRM), customer service management (CSM), business process management (BPM), and more.】. Microsoft and other companies have also joined the competition.
As AI agents appear in so many contexts and platforms, organizations interested in this technology may find it challenging to know where to start. According to AI experts, so far, a few use cases have emerged as frontrunners.
Consulting and tax service provider EY【Ernst & Young (EY) is a multinational professional services company headquartered in London, UK, and is one of the Big Four accounting firms. It has evolved through numerous mergers of various firms. Its organizational structure includes a management committee and multiple regions divided by geography, with over 700 offices in more than 150 countries and regions. Its services include audit, tax, consulting, and transaction advisory services for businesses, governments, and non-profit organizations, enjoying high recognition and influence globally, and has received numerous honors, ranking high in global brand value. EY also has extensive operations in China, assisting Chinese enterprises in their development.】‘s Global Innovation AI Officer Rodrigo Madanes【Rodrigo Madanes is EY’s Global Innovation AI Officer, leading AI innovation efforts and co-chairing the AI Center of Excellence, overseeing the AI startup ecosystem. He is dedicated to helping EY fully leverage AI’s potential by reshaping and transforming its services in accounting, tax, strategy, and consulting through technologies such as machine learning.】 stated that Agentic AI will seamlessly integrate with ERP【ERP stands for Enterprise Resource Planning, an information system that integrates and manages various internal resources of an enterprise (including human, financial, material, production, sales, etc.). By standardizing internal processes and centralizing data processing, it helps enterprises optimize resource allocation, improve operational efficiency, and enhance management levels.】, CRM【CRM, or Customer Relationship Management, focuses on interactions between businesses and customers. It is a customer-centric business strategy and management software that helps businesses understand customer needs better, improve customer satisfaction and loyalty, thereby increasing sales and profits.】, and business intelligence systems to achieve workflow automation, data management, and generate valuable reports. Unlike some past automation technologies, AI agents can make real-time decisions, making process automation a primary use case.
Madanes stated: “AI agents can automatically perform repetitive tasks that previously required human intervention, such as customer service, supply chain management, and IT operations. This technology’s uniqueness lies in its ability to adapt to changing conditions and handle unexpected inputs without human oversight.”
Here are six major use cases for AI agents as seen by several AI experts.
Agentic AI promises to transform AI coding assistants or copilots into smarter software development tools capable of writing substantial amounts of code. Although coding assistants have received mixed reviews so far, research firm Gartner【Gartner is a globally recognized information technology research and consulting firm that provides research reports, consulting services, and various tools for businesses in areas such as IT strategy, digital business, supply chain, and more. Its clients include businesses of all sizes, government agencies, educational institutions, and others.】 predicts that smarter AI agents will write most of the code within three years, leading most software engineers to need to relearn their skills.
Digital transformation consulting firm Publicis Sapient【Publicis Sapient is a globally recognized digital transformation consulting firm, including strategy and consulting, customer experience and design, technology and engineering, marketing platforms, data and AI, innovation and digital product management.】‘s Executive Vice President and Chief Product Officer Sheldon Monteiro【Sheldon Monteiro is the Chief Product Officer at Publicis Sapient, a global digital transformation consulting firm. He joined the company in 1995 and has been dedicated to leading teams to provide digital transformation services to businesses.】 stated that coding agents will not only write code but also have separate agents to check for errors in the code.
“As DevOps【DevOps is a set of processes, methods, and systems that integrate software development (Development) and IT operations (Operations), aiming to shorten system development cycles, improve software quality, and enhance service reliability. In simple terms, it breaks down the barriers between traditional software development and operations, promoting collaboration and communication between development and operations teams.】 tools have already automated workflows, adding AI agents is a natural evolution,” he said. “These agents can autonomously reverse-engineer specifications from code, forward-engineer test cases and code from specifications, and approve artifacts that meet certain threshold criteria, thereby improving overall automation levels.”
Many organizations are already using robotic process automation to automate simple repetitive tasks in various areas. Monteiro stated that Agentic AI can also automate tasks, but it can tackle more complex problems requiring higher-level decision-making capabilities.
“With AI, RPA【RPA stands for Robotic Process Automation, a technology that automates repetitive, rule-based tasks by simulating human interactions with computer systems through software robots. These software robots can operate applications like humans according to pre-defined rules, such as logging into systems, copying and pasting data, filling out forms, sending emails, etc.】 is no longer just rule-based action, but rather adaptive and autonomous, significantly improving overall business operation efficiency,” he said. “New tools allow us to train agents to not only complete the simplest tasks that RPA is currently performing but also truly understand the nuances of when exception logic comes into play.”
3. Customer Support Automation
AI customer experience solution provider Genesys【Genesys is a leading provider of customer experience and contact center solutions globally, offering multichannel customer experience and contact center solutions for businesses of all sizes, including products like Genesys Multicloud CX, Genesys Cloud CX, PureConnect, and Genesys DX, deployable in various environments.】‘s Chief Technology Officer Glenn Nethercutt【Glenn Nethercutt is the Chief Technology Officer and Technical Researcher at Genesys, responsible for overseeing cloud architecture strategy, including scalability, microservices, cloud-native design, fault tolerance, disaster recovery, service consistency, new technology assessment, and continuous delivery mechanisms.】 stated that businesses have long used simple chatbots and voice bots to handle basic customer service requests, but agent-based AI will evolve customer service automation into a more powerful service, rather than just answering some frequently asked questions.
“I tend to define agent-based AI as the ability to autonomously execute reasoning, multi-step, and non-deterministic tasks, a capability to handle genuinely complex and adaptive decision-making processes without human guidance,” Nethercutt said.
He mentioned that these AI customer service agents will cover various industries and functions, including retail, financial services, and IT service desk assistance. AI customer service agents will be able to understand and provide contextual answers to a wide range of customer needs, rather than just answering limited questions like a robot.
For example, a bank customer might say, “Withdraw money from my account with the most funds and transfer it to my checking account.” Nethercutt said simple chatbots typically cannot understand what “the account with the most funds” means.
“Our idea is to have a directory of actionable operations and an AI that is smart enough to navigate through the many options in front of me, where the guardrails will become increasingly complex,” he said.
Experts say that as ServiceNow【ServiceNow, founded in June 2004 and headquartered in Santa Clara, California, is a company providing enterprise cloud computing solutions. Initially focused on IT service management (ITSM), it provides cloud-based solutions for automating IT service requests, incident management, problem resolution, and configuration management, later expanding its product range to include a broader range of enterprise services such as human resources, customer service management, and security operations.】, Salesforce【Salesforce is a globally recognized provider of customer relationship management (CRM) software, occupying an important position in the CRM field.】 Its philosophy is to help businesses enhance sales, marketing, and customer service efficiency by providing powerful CRM solutions.】 and other vendors adopt agent AI, business workflows will become the best application area for this technology, enabling businesses to streamline processes by automating daily tasks.
For example, Monteiro said AI agents could convert meeting notes into project tickets without human input or trigger vendor orders based on supply and demand forecasts.
He added that organizations deploying large vendor IT tools across the enterprise should have an advantage over companies that may require various solutions linked through APIs. It is crucial for enterprises to consolidate all data and avoid information silos.
“The challenge for CIOs is, ‘Who do you want to delegate to build your context storage, i.e., your deep understanding of how the enterprise operates?’” he added. “Think about all you know about the enterprise. What if your JD really knew everything about how your enterprise operates?”
5. Cybersecurity and Threat Detection
Several cybersecurity providers have deployed AI agents to detect and respond to threats. Monteiro stated: “Agent AI in cybersecurity can autonomously detect, respond to, and even mitigate security and fraud threats almost in real-time, thereby reducing response times to potential attacks and enhancing overall security.”
Additionally, according to AI agent vendor Beam【Beam focuses on providing intelligent agent-related services utilizing AI technology. Such companies typically conduct in-depth research and applications in natural language processing, machine learning, and other fields, providing organizations with agents capable of simulating human behavior and intelligent decision-making.】, AI agents can enable personalized security protocols to adapt to specific threats and vulnerabilities. The company claims: “This agent automation ensures a stronger defense mechanism.”
Beam also states that AI agents can improve efficiency and cut costs by automatically performing routine tasks and security responses.
Another area where AI agents will have a significant impact is business intelligence. AI-driven BI【BI stands for Business Intelligence, which combines business analytics, data mining, data visualization, data tools and infrastructure, and best practices to help organizations make more data-driven decisions.】 vendors Zenlytic【Zenlytic is an AI-driven business intelligence platform】‘s co-founder and CEO Ryan Janssen【Ryan Janssen is the co-founder and CEO of Zenlytic, an AI-driven business intelligence platform, leading the team to apply AI technology in the field of business intelligence, driving companies to utilize data more efficiently for decision-making.】 believes that traditional BI dashboards often rely on data teams to gain deep insights, while combining AI with BI solutions can empower more employees to access useful analytics.
He stated that while BI dashboards are relatively simple to use, gaining insights beyond standard categories usually requires the work of data teams to extract.
He mentioned that AI agents combined with BI solutions can provide useful analytical data to more employees. For instance, AI agents in BI can provide marketing teams with suggestions on how to spend their budget or create charts based on examples drawn on napkins, Janssen said.
AI agents that understand voice input can generate business data insights based on verbal questions, such as “What are our top three marketing channels?”
“This is a very natural question, but it is vague,” Janssen said. “The problem that chatbots cannot solve compared to agents is eliminating this ambiguity. What do you mean by ‘top’? If the agent is built correctly, it will say, ‘Oh wait, this is ambiguous; I need to go back and use tools to clarify this.’”
Janssen added that many organizations are just beginning to use agent-based AI, and there are hundreds of potential use cases yet to be discovered. Coding agents are an early use case because programming is detail-oriented and time-consuming, but now coding enthusiasts are using coding assistants to build applications.
Janssen stated: “They perform best when your work is tedious, requires a lot of effort, or is very detail-oriented.”
He added that when dozens of agents are linked and organized, businesses will see new breakthroughs.
“We haven’t even scratched the surface of what agents can do,” he said. “We don’t yet know what organizations will look like, how they should interact, and how to manage them. But I have no doubt we will figure it out in the coming years.”
Authors: Jason Snyder and Grant Gross
【Insight: Agentic AI, a new type of artificial intelligence with autonomy, is revolutionizing various industries. It can not only execute preset tasks but also think and make decisions proactively like a human, significantly enhancing work efficiency and decision-making quality.
Branch Argument 1: Definition and Development of Agentic AI
-
Agentic AI: Artificial intelligence with autonomy, capable of perceiving the environment, making decisions, and taking actions.
-
Current Development Status:
-
Forrester has rated it as one of the hottest emerging technologies for 2025.
-
Several well-known companies (such as Salesforce, ServiceNow, EY) have begun exploring its applications.
Branch Argument 2: Application Scenarios of Agentic AI
-
Software Development:
-
Enhanced RPA:
-
Customer Support:
-
Business Workflows:
-
Cybersecurity:
-
Business Intelligence:
Branch Argument 3: Future Development of Agentic AI
-
Integration with Existing Systems:
-
Seamlessly integrate with ERP, CRM, and other systems.
-
Achieve more comprehensive automation.
-
New Application Scenarios:
-
Organizational Structure Changes:
Conclusion:
The emergence of Agentic AI marks a new stage in the development of artificial intelligence, which will have a profound impact on various industries. Businesses need to actively embrace this new technology to maintain a competitive advantage.
Summary Points:
-
Agentic AI is an artificial intelligence with autonomous learning and decision-making capabilities.
-
Agentic AI has broad application prospects in software development, RPA, customer support, business workflows, cybersecurity, and business intelligence.
-
Agentic AI will change the way enterprises operate and have a profound impact on human work.
Recommendations:
-
Businesses should closely monitor the development trends of Agentic AI and formulate corresponding strategies.
-
Organizations should cultivate employees’ ability to adapt to the changes brought by Agentic AI.
-
Governments should formulate relevant policies to promote the healthy development of Agentic AI.
