Wanted: First movers for Southeast Asia’s agentic AI leap

If you say you know what agentic AI is, you’re probably lying.

Sure, you might have a few platitudes that you fire off to sound informed. But industry insiders tell Tech in Asia that when they speak at conferences, including those here in Southeast Asia, it’s clear that most executives wouldn’t know an AI agent from a large language model (LLM).

“Chatbots, AI, or agentic AI – they have no idea what the difference is,” says Jason Sosa, CEO of Element1, a Singapore-based software firm specializing in AI. “I think that is the biggest pain point that we find.”

Because executives don’t understand the nuances of AI, they don’t know what to ask for when talking to developers, according to Sosa.

Image credit: Timmy Loen

So what is agentic AI? For the record, it’s text to action instead of just text to text, Sosa explains.

For companies, agentic AI means automating large parts of their organization, such offering customer service rebates or designing an app that could schedule all your appointments for you.

See also: SEA’s AI players target end users as DeepSeek stifles LLMs

Agentic AI has the potential to change how humans interact with everything.

Users would no longer need to open an app, search the web, or speak to a person. Instead, they could just pick up their phones and tell their AI agent to schedule a holiday. The agents help users figure out where to go, who to fly with, and where to stay.

Keep your cards in your wallet. In the end, your AI agent can even handle the payment for you.

The global market value of agentic AI was estimated at US$5.2 billion in 2024 and is expected to hit US$197 billion by 2034.

Agent of chaos?

Beware of strange bots

Not all AI bots are agentic

But can you trust it?