The AI world is never in want of news. Whether it’s about DeepSeek’s disruption of the global AI space or the U.S. announcing Stargate — a $500 billion private sector deal to expand U.S. AI infrastructure — there’s something always in the works. While generative AI and artificial general intelligence have dominated headlines over the last two years, the new phrase on the lips of many players in the AI space is agentic AI — intelligent systems or so-called “agents” that don’t just generate content but autonomously reason, adapt, and take action.
One report by Deloitte predicts that “25% of companies that use generative AI will launch agentic AI pilots or proofs of concept in 2025, growing up to 50% by 2027.” The report further notes that some agentic AI applications, in some industries, and for some use cases, could see actual adoption into existing workflows in 2025, especially by the back half of the year.
But what’s fueling this shift and where do we go from here?
The Tipping Point: Why 2025 Is The Year Of Agentic AI
Generative AI models like GPT-4 and Claude Sonnet 3.5 have certainly transformed the way businesses interact with data. Tools like ChatGPT and Gemini have changed content creation, automation, and customer interactions. However, these models are limited by their inability to learn from or integrate with your unique business data. And that’s where agentic AI comes in.
Unlike traditional generative AI, which produces text or images in response to prompts, agentic AI operates with autonomy, making decisions based on real-time data. It doesn’t just generate — it acts. Ken Mahoney, CEO of Mahoney Asset Management, said that “agentic AI essentially is an even more advanced version of generative AI.”
While generative AI is good at instant research, generating text and images and is overall a tool for human assistance in general, he explained that agentic AI moves beyond some of these simple tasks and demands AI to actually complete higher level tasks like real decision making and problem solving, and acts as a consultant, or helps act on behalf of businesses or individuals making higher level decisions.
“Agentic AI bridges the gap between AI’s potential and business reality,” said Vahagn Sargsyan, founder and CEO of WebWork AI. “We’ve seen AI transform industries, but companies need AI that understands their data, not just general internet knowledge.” At WebWork, for example, this evolution is already happening. The company’s AI assistant analyzes tracked time data to provide actionable insights on productivity, workflow inefficiencies, and operational anomalies — helping businesses optimize processes and scale intelligently.
Kevin Frechette, cofounder and CEO of Fairmarkit, expanded on this idea, emphasizing the move from generative AI to agentic networks. “Agentic AI means agentic networks — teams of AI agents collaborating and even having supervisor agents oversee their workflows. In procurement, for instance, AI can now negotiate terms, verify compliance, and manage vendor interactions autonomously, significantly reducing the burden on human teams,” he told me in an interview.
Smart Assistants, Not Replacements
Like humans, AI also makes mistakes. A notable example of such a mistake was when tutoring company iTutor Group’s recruiting AI rejected applicants based on their age — a conduct which violated the Age Discrimination In Employment Act. The company eventually paid a settlement of $365,000 in a lawsuit filed by the U.S. Equal Employment Opportunity Commission (EEOC).
That’s why, for Sargsyan, AI should augment human capabilities rather than replace them. “AI should be a human assistant, not a replacement,” he explained. “At WebWork AI, our system doesn’t just monitor data — it acts as an assistant, analyzing massive datasets and surfacing the insights that truly matter. This allows businesses to make informed decisions without drowning in data.”
Frechette echoed Sargsyan’s sentiment, noting that it’s important to prioritize bias detection and transparency in AI systems. “The future of AI isn’t about blindly trusting automation. It’s about designing AI-driven workflows where humans stay in control. AI ethics oversight will become a major focus area for companies looking to deploy agentic AI responsibly,” he said.
Business Applications of Agentic AI
There are real questions being asked about whether AI agents could truly unlock efficiency and scalability. But experts like Nadja Atwal, startup advisor and member of the Global AI Council Leader Board, believe that companies are driving innovation in agentic AI by rapidly prototyping and scaling solutions that solve complex problems. There’s a broad range of use cases for AI agents, from healthcare to finance and more.
In healthcare, for example, AI agents can assist with patient triage, diagnostics and personalized treatment plans, expanding access to care. In supply chain and logistics, Atwal said that AI agents can “dynamically optimize routes, manage inventory, and predict demand, reducing waste and costs.”
When we zoom in on finance, said Atwal, agentic AI is great for autonomous trading, fraud detection and personalized financial advisory services. Finq.ai is already leveraging AI agents in this manner. According to Eldad Tamir, CEO of Finq.ai, the company’s stocks technology tool uses AI to analyze market trends, rank stocks and generate model portfolios that consistently outperform the S&P 500.
Adapting To Regulations
While AI agents offer great promises that could truly shape how businesses operate globally, they also open a can of worms when it comes to regulations. The EU, for example, has ramped up AI governance frameworks, demanding higher transparency and accountability, per a report by Reuters.
This is why Tamir believes ethical AI is going to be crucial in the adaptability of AI. “Regulators are still very much confused on what and how they should implement regulation. At this stage, it’s up to executives to take full responsibility and make good and ethical use of AI, which begins with a commitment to transparency and accountability. Businesses must ensure compliance with regulations by building systems that prioritize data privacy, fairness and explainability,” he said.
For Frechette, companies can “expect to see more interdisciplinary AI ethics boards and regulatory engagement to ensure AI systems are fair, unbiased and accountable.”
A Difference Maker
Tamir argued that this is a turning point in history, particularly because “for the first time ever, machines will take part in actual decisions within corporations, as businesses move beyond experimentation into practical implementation.”
While the full impact of agentic AI is still unclear, there’s no doubt that more businesses will jump on this train as the year unfolds. As Mahoney noted, companies like Walmart are already using agentic AI to their advantage — to autonomously manage inventory, predict demand for certain products and offer real time recommendations that are targeted purposefully to each customer.
“This is going to be the difference maker for companies who can properly adopt agentic AI and implement it in their businesses to find efficiencies like Walmart has,” he said.
