“`html
AI agents are revolutionizing the business world. A new study from Perplexity reveals how these autonomous tools are taking over complex enterprise tasks, radically transforming workflow efficiency. Who is using these technologies? How are professionals integrating them into their daily work? What are the real use cases in the enterprise?
The era of agentic AI arrives in enterprises
Over the past year, the technology sector has anticipated the next evolution of generative artificial intelligence: the shift from conversation to action. While large language models serve as reasoning engines, agents act as the hands capable of executing complex multi-step workflows with minimal supervision.
Until now, visibility on the actual use of these tools remained limited, relying primarily on speculative frameworks or restricted surveys. New data published by Perplexity, analyzing hundreds of millions of interactions with its Comet browser and assistant, provides the first large-scale field study on general-purpose AI agents.
The results indicate that agentic AI is already being deployed by high-value knowledge workers to streamline productivity and research tasks.
Who is adopting AI agents? User profiles
Understanding who uses these tools is essential for anticipating internal demand and identifying potential “shadow IT” vectors. The study reveals marked heterogeneity in adoption. Users in countries with higher GDP per capita and education levels are much more likely to use agentic tools.
More revealing for enterprise planning is the professional breakdown. Adoption is heavily concentrated in digital and knowledge-intensive sectors:
- The “Digital Technology” cluster represents the largest share with 28% of users and 30% of queries
- Followed closely by academia, finance, marketing, and entrepreneurship
- Collectively, these clusters account for more than 70% of total users
This concentration suggests that those most likely to leverage agentic workflows are an organization’s most expensive assets: software engineers, financial analysts, and marketing strategists. These early adopters are not just testing the technology: the data shows that “power users” perform nine times more agentic queries than average users, indicating that once integrated into a workflow, the tool becomes indispensable.
Cognitive assistants, not digital butlers
To go beyond marketing discourse, enterprises must understand the real utility of these agents. A common vision suggests agents will function primarily as “digital concierges” for routine administrative tasks. However, the data contradicts this view: 57% of all agent activity is concentrated on cognitive work.
Perplexity researchers developed a “hierarchical agentic taxonomy“ to classify user intent, revealing that the use of AI agents is practical rather than experimental:
- Productivity and workflow dominate with 36% of all agentic queries
- Learning and research follow with 21%
Specific anecdotes from the study illustrate how this translates to enterprise value. A procurement professional, for example, used the assistant to scan customer case studies and identify relevant use cases before engaging with a supplier. Similarly, a finance worker delegated tasks of filtering stock options and analyzing investment information.
This distribution provides clear guidance to operational managers: the immediate ROI of agentic AI lies in augmenting human capacity rather than simply automating low-level friction.
User stickiness and cognitive migration
A key insight for IT managers is the “stickiness” of AI agents to enterprise workflows. The data shows that in the short term, users exhibit strong persistence within the same subject. If a user engages an agent for a productivity task, subsequent requests will likely remain in that domain.
However, the user journey often evolves. New users frequently “test the waters” with low-stakes queries, such as asking for movie recommendations or general questions. Over time, a transition occurs. The study notes that while users may enter through various use cases, query shares tend to migrate toward cognitively-oriented domains such as productivity, learning, and career development.
Once a user employs an agent to debug code or summarize a financial report, they rarely return to lower-value tasks. The “Productivity” and “Workflow” categories demonstrate the highest retention rates.
The preferred environments for AI agents
The “where” of agentic AI is just as important as the “what.” The Perplexity study tracked the environments—specific websites and platforms—where these AI agents operate. The concentration of activity varies by task, but the main environments are pillars of the modern enterprise technology stack:
- Google Docs is a primary environment for document and spreadsheet editing
- LinkedIn dominates professional networking tasks
- For learning and research, activity is divided between learning platforms like Coursera and research repositories
For CISOs and compliance managers, this presents a new risk profile. AI agents don’t just read data; they actively manipulate it within core enterprise applications. The study explicitly defines agentic queries as those involving “browser control” or actions on external applications via APIs.
The concentration of environments also highlights the potential for platform-specific optimizations. For example, the five leading environments represent 96% of professional networking queries, primarily on LinkedIn. This heavy concentration suggests that enterprises could see immediate efficiency gains by developing governance policies tailored to these high-traffic platforms or API connectors for them.
Strategic planning for agentic AI in the enterprise
The proliferation of capable AI agents invites new thinking for enterprise planning. Perplexity’s data confirms we have moved beyond the speculative phase. Agents are currently being used to plan and execute multi-step actions, modifying their environments rather than simply exchanging information.
Operational managers should consider three immediate actions:
Audit friction points in productivity and workflow
The data shows this is where agents naturally find their place. If software engineers and financial analysts are already using these tools to edit documents or manage accounts, formalizing these workflows could standardize efficiency gains.
Prepare for the reality of augmentation
Researchers note that while agents have autonomy, users often decompose tasks into smaller elements, delegating only sub-tasks. This suggests that the immediate future of work is collaborative, requiring employees to upskill on how to effectively “manage” their AI counterparts.
Strengthen the infrastructure and security layer
With agents operating in “open web environments” and interacting with sites like GitHub and corporate email, the data loss prevention perimeter expands. Policies must distinguish between a chatbot offering advice and an agent executing code or sending messages.
As the agentic AI market is expected to grow from $8 billion in 2025 to $199 billion in 2034, Perplexity’s early evidence serves as an indicator. The transition toward agent-driven enterprise workflows is underway, driven by the most digitally capable segments of the workforce.
The challenge for the enterprise is to leverage this momentum without losing the governance control necessary to scale it safely.
Source: Artificial Intelligence News
“`
