In the theater of the 21st century, the stage is not set by human actors alone. We have entered the era of multi-agent generative AI systems, a shift in which artificial intelligence moves beyond discrete tools such as search engines or calculators to become distributed, interacting systems embedded in commercial and institutional ecosystems. While algorithmic nudging has shaped our everyday behavior for years, from product recommendations on Amazon to content curation on Netflix, a more consequential transition is now underway. We are witnessing the rise of algorithmic swarms, disconnected yet invisibly linked like bees in a hive.

As AI agents increasingly operate not in isolation but in coordinated networks, their role extends beyond task automation to large-scale influence over information flows, attention, and decision environments. Moving beyond single programs that respond to individual users, this is about many interconnected digital entities—mostly optimized for engagement, efficiency, and profit—interacting across platforms and contexts. Together, they contribute to emergent patterns that can subtly steer collective behavior and shape public discourse at scale. What is changing is not simply the sophistication of individual systems, but the systemic nature of their interaction, and the societal consequences that follow.

To grasp this shift, one must distinguish between standard AI and agentic AI. While standard AI waits for a prompt to provide an answer, agentic AI is designed to have agency, the ability to set its own sub-goals, use digital tools, and interact with........

© Psychology Today