These top 30 AI agents deliver a mix of functions and autonomy

Source: ZDNet

ZDNET’s key takeaways

• MIT released a list of top AI agents and their functionalities.
• The largest portion focuses on enterprise workflows.
• Research and information synthesis is the top use case.

Which autonomous or semi‑autonomous agents are making the greatest impact on the world — and potentially your job — these days? Certain agents dominate the headlines, but a variety of function‑specific agents are available to developers and users.

The researchers found that interfaces are the most abundant, followed closely by enterprise workflow platforms. They also uncovered risks shared across these agents, as explored by Tiernan Ray in a related ZDNet article.

Agents featured in the MIT index

• Anthropic Claude / Claude Code
• Google Gemini / Gemini CLI
• Manus AI
• OpenAI ChatGPT / ChatGPT Agent / Codex / AgentKit
• Perplexity
• Alibaba MobileAgent
• ByteDance Agent TARS
• Perplexity Comet
• IBM watsonx Orchestrate
• Microsoft 365 Copilot
• SAP Joule Studio
• Salesforce Agentforce
• ServiceNow AI Agents

Agent categories identified by the researchers

Enterprise workflow agents (13 of the 30 systems)

Platforms with agentic features for automating business tasks.
Examples: Microsoft 365 Copilot, ServiceNow Agent.

Chat applications with agentic tools (12 systems)

Primarily chat interfaces with extensive tool access.
Examples: Claude Code, Manus AI, ChatGPT Agent.

Browser‑based agents (5 agents)

Agents whose primary interface is the browser or computer, with extensive interaction tools.
Examples: Perplexity Comet, ChatGPT Atlas, ByteDance Agent TARS.

“Browser‑based agents present higher risks through background execution, event triggers, and direct transactions.” – MIT study

Chrome’s AI agent shop, research, and email for me – here’s how it went

What are the most popular uses for AI agents?

Top use cases, cutting across the categories, include:

• Research and information synthesis – featured in 12 of the 30 agents, spanning consumer chat assistants and enterprise platforms.
• Workflow automation – present in 11 agents, primarily enterprise products, covering HR, sales, support, and IT functions.
• GUI or browser capabilities – used for tasks such as forms, ordering, and booking, found in seven models.

6 reasons why autonomous enterprises are still more a vision than reality

Levels of autonomy

• Chat‑first assistants (lowest autonomy): Turn‑based interactions that execute a single set of actions and wait for the next user prompt. Examples: Anthropic Claude, Google Gemini, OpenAI ChatGPT.
• Browser agents (higher autonomy): Limited opportunities for mid‑execution intervention; once a query is sent, the agent runs to completion without easy user steering. Example: Perplexity’s Comet.
• Enterprise platforms (variable autonomy): Users configure triggers, actions, and guardrails during design; once deployed, many operate autonomously, triggered by events such as new emails or database changes. Examples: Glean, Google Gemini Enterprise, IBM watsonx, Microsoft 365 Copilot, n8n, OpenAI AgentKit.

Google Workspace now lets you create AI agents to automate your work – how to get started

Developer / CLI agents

A few offerings require explicit confirmation for sensitive operations (e.g., file edits, command execution). Some provide a “watch mode” for real‑time oversight of critical actions, such as ChatGPT Agent/Atlas and Opera Neon.

The study also noted that agent developers are concentrated in the United States and China, with limited representation from other regions.