Summary Bullets:

• At this year’s AWS Summit in Sydney, AI has moved further along from conversation around large language models (LLMs) to practical AI examples driven by agentic AI.

• AI adoption is now across start-ups, SaaS companies, service providers, and enterprises with modern tech stacks.

It is no surprise that at the AWS Summit Sydney 2025, AI is still front and center. What is different from a year ago however, is the shift in conversation from LLMs and what they can potentially achieve to more practical AI including agentic AI and how they are now being deployed across different scenarios. While AI companies continue to formulate new models and enhance their accuracy and capability, there are now more ways enterprises can extract the value from AI models for different use cases. When posed a question how enterprises should go about selecting the right AI model, an AI advisory firm explained aptly that it should depend on the problem AI is meant to solve, and the level of IQ you need for that purpose. A less sophisticated, but more economical AI model, should suffice if the purpose does not require extensive reasoning or an LLM that is capable of performing a wide range of tasks.