Abstract

The emergence of Agentic AI will mean that artificial intelligence is stepping out of the phase of machines that work under commands given by human beings and will start functioning as independent, goal-oriented intelligent systems. AI workers, which are specialised computer agents, work together to deal with different parts of people’s professional and personal lives in a systematic way.

Agentic AI is a breakthrough in the development of artificial intelligence, shifting from mere response systems to autonomous agents that plan, reason, decide, and act for the purpose of accomplishing certain objectives. With its combination of advanced language modelling technology, tools, memory, and decision-making capabilities, agentic AI can revolutionise various industries, including medicine, banking, education, manufacturing, and customer service, among others. This talk examines the role of agentic AI with multi-agent capabilities, the opportunities it brings in terms of increasing productivity and innovation, as well as the issues related to reliability, accountability, security, and governance of AI. As machines learn to function with minimal human interaction, the questions of designing, deploying, and regulating agentic AI will acquire special importance. The future of artificial intelligence could thus be shaped not only by what machines understand and produce but also by what they can do on their own.