Editorial illustration for Analytical AI predicts outcomes but isn’t agentic, experts say
Analytical AI predicts outcomes but isn’t agentic,...
Analytical AI predicts outcomes but isn’t agentic, experts say
The buzz around AI agents has gone from niche to mainstream in just a few months. A November 2025 report from MIT Sloan School of Management and Boston Consulting Group shows 35 percent of surveyed firms already run AI agents, while another 44 percent say they’ll roll them out soon. That rapid uptake prompted MIT News to sit down with Phillip Isola, an associate professor in EECS and a CSAIL researcher, to unpack what “agentic AI” really means.
Isola draws a clear line between the new agents and the generative models most people know, like ChatGPT or Claude. He says agentic AI is defined by action—whether a robot picks up an object or a bot books a flight—whereas generative AI simply produces text, images or music. The term “agent” is largely a branding choice, often referring to software that helps users navigate a website, an app, or even a physical environment.
Today’s agents are mostly digital, such as the chat assistants that handle product complaints. Behind the scenes, many companies rely on a handful of the same underlying models, adding the ability to act and retain context.
Analytical AI methods, like the systems that help predict possible outcomes of decisions, are not agentic in nature, but are very informative to human decision-makers. For cases that are either high-stakes or safety-critical, like medicine, security, high-level business policies, etc., the technology might not be ready for AI to completely automate those processes, or we might not even be comfortable with that.
Q: Are there risks we should be thinking about when using AI agents?
A: One big risk area comes from the fact that it is often very easy to get agents to do certain types of work for you.
Why this matters We see a surge: 35 % of firms already run AI agents, another 44 % plan to follow, according to the MIT Sloan‑BCG report. Yet the analytical tools that forecast outcomes are not agentic; they serve as decision‑support, not autonomous actors. For high‑stakes domains—medicine, security, corporate policy—this distinction matters.
Developers must remember that predictive models can inform but cannot replace human judgment where safety is critical. Founders should weigh the allure of “agentic” hype against the proven utility of analytical AI, especially when regulatory scrutiny looms. Researchers are left with a clear research gap: how to integrate informative analytics into agentic frameworks without compromising oversight.
Unclear whether the rapid rollout of agents will translate into measurable improvements in those high‑risk sectors, or simply add layers of complexity. Our takeaway: adopt analytical AI where it adds clarity, but stay vigilant about the limits of agency and the need for human control.
Further Reading
- Agentic AI, explained | MIT Sloan - MIT Sloan
- Beyond Analysis: How Agentic AI Turns Data into Actions - Canton Group
- What is Agentic AI? - IBM - IBM
- AI Agents: What They Are and Their Business Impact | BCG - Boston Consulting Group
- What is Agentic AI and How It Works: 8 Real-World Use Cases - Exabeam
Related Reading
Market Trends 60% of singles label AI relationships as cheating, sparking divorce concerns
Market Trends Rajan says AI-native fundamentals let new grads outpace senior devs
Market Trends Aging grid threatens OpenAI, Microsoft as hyperscalers commit USD 400B to centers
From Archives Internet, cloud, and big data drive AI into large‑model era, but use stalls
From Archives