10
F
or decades, AI in the
workplace (especially
in legal) functioned
mainly as a tool, an
analytical engine, or an efficiency
booster directed by humans.
Today, we are witnessing a leap
toward agentic AI: systems that
can plan, decide, and act with
minimal oversight, even guiding
human coworkers (https://hbr.
org/2024/12/what-is-agentic-ai-
and-how-will-it-change-work). An
AI agent is an autonomous system
that can perceive its environment,
make decisions, and take actions
to achieve specific goals without
constant human direction. Unlike
traditional AI tools that respond
to direct commands, agents can
independently plan multi-step
processes, adapt to changing
circumstances, and execute
complex tasks from start to finish.
These agentic "co-workers"
don't just answer questions or
automate single tasks; they can
manage multi-step workflows
and make independent
decisions. For example, an AI
agent might autonomously
debug a software application by
identifying errors, researching
solutions, implementing fixes,
and running tests to verify the
repairs. Alternatively, in legal
contexts, it could continuously
monitor regulatory changes and
automatically flag contracts that
require updates, while drafting
compliance memos for review and
approval. Such capabilities mark a
shift from AI as a passive assistant
to AI as an active collaborator with
a degree of ownership over tasks.
WHEN AI
DELEGATES
TASKS TO
YOU
FEATURES
BY ABHIJAT SARASWAT
