Looking back, 2023 and 2024 will be remembered as the "Years of Amazement" with Generative AI. We were dazzled watching ChatGPT write poems in seconds or Midjourney create surreal paintings from a simple text prompt. The technology was extraordinary, but at heart it was still passive. It waited for our command, produced something, then stopped.
At the end of 2025, though, the technology landscape has shifted fundamentally. Amazement has turned into a need for more concrete usefulness. We no longer just need AI that is good with words; we need AI that is good at work.
This is the dawn of a new era: the rise of Agentic AI (autonomous artificial intelligence).
What Is Agentic AI?
To understand Agentic AI, we need to distinguish it from its predecessor, Generative AI.
- Generative AI (the past): like a very talented artist or writer sitting quietly in a room, waiting for specific instructions. "Write a draft email to the client," you say. It writes it perfectly, then falls silent again, waiting for you to copy, paste and send it.
- Agentic AI (the present): like a proactive executive assistant. You give an abstract goal: "Please handle the contract renewal with Client X." The agent doesn't just write an email. It will (1) open your calendar to find free slots, (2) open the old contract to check its terms, (3) draft a renewal proposal, (4) email the client and (5) schedule a meeting if the client agrees, all autonomously, without you directing each step.
The essence of Agentic AI is the shift from answering questions to completing tasks. Agents can plan, use other digital tools, make decisions based on the situation and carry out a series of complex actions to reach a bigger goal.
Real Impact on Indonesia's Business Landscape
The move to Agentic AI is not just theory in Silicon Valley labs; by the end of 2025 its impact is already significant in Indonesia's business ecosystem. Local companies are realising that efficiency is no longer about typing faster but about automating entire workflows.
Here are two main areas where Agentic AI is bringing big changes to the country:
1. Business Automation That Is "Smart", Not Rigid
Earlier automation (such as Robotic Process Automation, or RPA) was very rigid. If A happens, do B. If the data deviated even slightly, the system got stuck.
Agentic AI changes the game. In Indonesia's busy e-commerce and logistics sector, for example, intelligent agents are now used for supply chain management. An AI agent can monitor warehouse stock in Jakarta and Surabaya in real time.
When stock runs low, the agent doesn't just send a notification. It independently compares prices from three suppliers, weighs shipping cost and time, picks the best option and even raises a draft purchase order for a human manager to approve. Hours of admin shrink to minutes.
2. Customer Service That Feels Genuinely "Human"
For years we were frustrated by customer-service chatbots that only offered FAQ articles or went round in circles without a solution. That was the old generation of AI.
At the end of 2025, major telecom and banking companies in Indonesia began launching customer service built on Agentic AI. The difference is striking. These agents don't just "understand" your complaint; they have the authority to "act" in the company's systems.
Example scenario: a customer complains that their bill has jumped.
- Old chatbot: "We apologise for the inconvenience. Please check your usage details in the app or call our call centre on 123." (Leaving the customer even more frustrated.)
- Agentic AI: "I can see a spike in roaming data use on the 15th. Were you abroad at the time? If this is a system error, I can submit a bill adjustment right now and temporarily block roaming to keep you safe. Shall I go ahead?"
The agent can access billing data, analyse anomalies, offer concrete solutions and carry out the fix in the backend system, an interaction that feels far more empathetic and helpful, like talking to a competent member of staff.
The Future: Collaboration, Not Replacement
The rise of Agentic AI naturally raises the classic worry: will it replace people?
At the end of 2025, the answer leans towards collaboration. Agentic AI takes over repetitive tasks that involve tedious navigation across applications, freeing people to focus on strategy, creativity and high-level decisions that need ethical and emotional nuance.
We are moving from an era where we use AI as a "tool" (like a hammer or a calculator) to one where we manage AI as "digital co-workers", or agents. The biggest challenge ahead is not making AI smarter, but learning, as humans, to be good managers of these autonomous agents.
