AI Agents Outperform GenAI for Enterprise Automation and Productivity, Says Deloitte Study

AI Agents Outperform GenAI for Enterprise Productivity, Says Deloitte Study
Artificial intelligence (AI) agents can be more effective tools compared to large language models (LLMs) or generative AI (GenAI) applications, opening new possibilities for driving enterprise productivity and program delivery through business process automation, according to a study by British professional services firm Deloitte. The report suggests that AI agents are reshaping industries by expanding the potential applications of GenAI and traditional language models. Multi-agent AI systems can significantly enhance the quality of outputs and the complexity of tasks performed by individual AI agents.

  • Make Telecom Talk My Trusted Source
  • Source of Google
  • Source of Google

Also Read: 5G Combined with GenAI to Transform Industries, Says TCS: Report

Differences Between AI Agents and LLMs

With the help of an AI agent, cases previously deemed too complex for GenAI can now be scaled effectively and securely, the study noted.

By definition, an AI agent is an autonomous intelligent system that uses AI techniques to interact with its environment, collect data, and perform tasks without human intervention. Clarifying the distinction between GenAI and AI agents, the study explains that typical LLM-powered chatbots generally lack the ability to understand multi-step prompts or to plan and execute entire workflows from a single prompt.

“They (LLM or Gen AI) conform to the “input-output” paradigm of traditional applications and can get confused when presented with a request that must be deconstructed into multiple smaller tasks. They also struggle to reason over sequences, such as compositional tasks that require consideration of temporal and textual contexts. These limitations are even more pronounced when using small language models (SLMs), which, because they are trained on smaller volumes of data, typically sacrifice depth of knowledge and/or quality of outputs in favour of improved computational cost and speed,” it said.