We’ve all been stuck in a never-ending customer service phone loop, trying to get a problem fixed with a product or service. It’s frustrating and time-consuming. Many wonder if technology has made it any better. When Gartner declared agentic artificial intelligence (AI) as a major tech trend for 2025, it felt a bit empty—like we’ve heard this all before. Can this technology really help automate decisions and support smarter AI or even humans?
Gartner describes agentic AI as a digital workforce that can operate independently, without needing breaks or benefits. The idea is that these AI agents can coordinate data and streamline processes, leading to a bold prediction: by 2029, they could resolve 80% of common customer service issues on their own, cutting operational costs by 30%. GlobalData echoes this idea, calling agentic AI a transformative force in automation.
Kiran Raj from GlobalData emphasizes that agentic AI could ease administrative burdens across various sectors, from healthcare to finance. Its ability to adapt in real-time is critical for businesses that need to make quick decisions. But here’s the catch: despite all the predictions, actual adoption of AI is slow. It’s mostly still in the testing phase rather than widespread use.
Nvidia CEO Jensen Huang recently called agentic AI “the next wave” of intelligent systems—agents that can think, plan, and act autonomously. Eirik Pettersen, CTO at Secret Escapes, has started testing these claims. His company implemented Salesforce’s Agentforce, an AI platform designed to support staff and customers continuously. Pettersen noted a drop in negative customer feedback after adopting the GenAI version of their chatbot, which shows real progress in enhancing the customer experience.
Forrester’s research highlights agentic AI as a game changer for businesses willing to invest strategically. Early adopters that align their efforts with business goals stand to unlock growth and efficiency. Yet, they also need to rethink their operational models and ensure strong AI foundations.
The distinction with agentic AI is its autonomy and ability to handle complex tasks independently. Forrester outlines a phased evolution of agentic AI—from basic task handling to multi-agent collaboration across organizations. Yet, not all companies are ready to fully trust AI. At Secret Escapes, they’re cautious, aiming to reduce reliance on human agents and manage costs carefully.
Empowering non-technical teams is crucial. Pettersen appreciates that his customer service head has been directly involved in developing AI prompts. This way, team members can improve solutions without always relying on technical staff. Early use cases at Secret Escapes demonstrate the flexibility of agentic AI, tackling everything from baggage inquiries to processing cancellations.
Other companies are diving in too. BT and OpenTable are experimenting with agentic systems to improve efficiency in customer support. There’s promise here, but as DeepMind’s CEO pointed out, AI is not infallible. Organizations need to ensure their data processes are solid to avoid compounding errors.
Consultants are now pushing businesses to explore how agentic AI can transform their operations. Simon James from Publicis Sapient points out that organizations need to decide how much autonomy they’re comfortable giving to AI while also ensuring human oversight.
Costs are also a concern. James notes that companies need to be aware of the rising expenses associated with cloud services for AI applications. Agentic AI shouldn’t just be another tech trend; it should change how businesses operate.
Investment in agentic AI is on the rise, with significant venture capital flowing into startups focusing on this tech. Yet, past investment spikes raise questions about sustainability. Companies must approach agentic AI thoughtfully, or risk losing sight of its true potential as an enhancer of human capabilities rather than just a cost-saving tool.
Ultimately, how organizations choose to use agentic AI—whether to innovate or just reinforce old ways—will be a key factor in its success.