Different Types of AI

Now that you've built a foundation in AI, machine learning, and automation, it's time to explore the three distinct types of AI you'll encounter most often in HR technology. Understanding these differences isn't just academic—it directly affects how you evaluate tools, set expectations, and communicate AI capabilities to stakeholders.

The three types we'll cover are generative AI, predictive AI, and agentic AI. Each serves a fundamentally different purpose: generative AI creates, predictive AI forecasts, and agentic AI acts. As you work through this unit, you'll develop the ability to quickly categorize any AI tool you encounter, which helps you ask sharper questions and make smarter purchasing decisions. By the end, you'll understand not just what each type does, but where it fits best in your HR workflows.

Generative AI refers to AI that creates new content—text, images, audio, video, or code—based on patterns learned from data. These systems use large models, often called Large Language Models or diffusion models, trained on massive datasets to produce outputs that didn't exist before. The defining characteristic is that generative AI is creative in nature—its core purpose is to produce something new.

You've likely already used generative AI tools like ChatGPT, Claude, or Microsoft CoPilot. When you ask ChatGPT to "draft an email declining a candidate respectfully" or "rewrite this job description to be more inclusive," you're leveraging generative AI. The system doesn't retrieve a pre-written template; it generates a fresh response based on your specific request and the patterns it learned during training.

For HR professionals, generative AI shines in content-heavy tasks. You might use it for writing inclusive job descriptions that attract diverse candidates or summarizing lengthy interview notes into actionable feedback. Furthermore, it proves valuable when drafting onboarding communications for new hires and creating training materials or policy documentation. The practical applications extend to everyday communications too—think of it as a capable writing partner who can produce first drafts, suggest alternative phrasings, or adapt content for different audiences.

The key judgment call with generative AI is knowing when its creative output needs human review. A job description draft might need adjustments for legal compliance, while an employee communication might require a tone check to ensure it reflects your company culture. Generative AI accelerates your work, but you remain the decision-maker who validates and refines the output.

Predictive AI refers to AI that forecasts outcomes or identifies patterns based on historical and real-time data. These systems use machine learning and statistical models to make predictions, classifications, or recommendations. Unlike generative AI, predictive AI is analytical in nature—its core purpose is to predict what might happen or classify what something is.

You encounter predictive AI constantly, even outside work. When Netflix recommends a show you might enjoy, that's predictive AI analyzing your viewing patterns. When your email filters spam, that's a predictive model classifying messages. In the HR context, predictive AI helps you anticipate outcomes rather than just react to them.

Consider the practical applications of predictive AI for your work:

• Forecasting turnover risk from engagement survey data and performance patterns, allowing you to intervene before resignations occur.
• Recommend training and development paths based on career trajectory analysis, helping employees grow in directions that align with organizational needs.
• Screening resumes to predict role fit and identify qualified candidates.
• Forecasting hiring needs based on business growth trends.

The conversation around predictive AI in HR often sounds like this: "Our people analytics vendor says their tool can predict which employees are flight risks. Should we trust it?" A thoughtful response might be "It depends on what data they're using and how accurate their predictions have been historically. What's their track record?" When the vendor reports 78% accuracy on turnover predictions within six months, that provides useful context—but the real question becomes what you do with those predictions. You need to ensure you're using the insights to support employees, not to make assumptions about individuals. This balance between technical validity and ethical application is essential when working with predictive AI.

Agentic AI refers to AI that can act autonomously toward a goal, chaining together multiple actions or tools without human intervention. These systems combine reasoning, planning, and execution—often using generative or predictive AI under the hood—to accomplish multi-step objectives. The distinguishing feature of agentic AI is that it's action-oriented—its core purpose is to decide and do.

While generative AI produces content and predictive AI provides insights, agentic AI takes things a step further by actually performing tasks end-to-end. Think of it as the difference between an AI that drafts an interview invitation email versus an AI that schedules the interview by checking calendars, finding available times, sending invites, and handling rescheduling requests—all without you stepping in.

For HR professionals, agentic AI represents the frontier of automation. Here are some current and emerging applications:

The promise of agentic AI is significant time savings on repetitive, multi-step tasks. However, this autonomy also requires careful setup. You need clear guardrails defining when the agent should escalate to a human, what actions it's authorized to take, and how to handle edge cases. As agentic AI matures, your role shifts from executing tasks yourself to designing workflows, setting boundaries, and monitoring outcomes.

Understanding these three types—generative for creating, predictive for forecasting, and agentic for acting—gives you a practical framework for evaluating any AI tool. Let's see how this framework works in practice through a conversation between two HR colleagues evaluating new technology.

• Jessica: I've been looking at three different HR tools this week and I'm trying to make sense of what they actually do.
• Dan: What are they claiming?
• Jessica: One writes job postings, another predicts which new hires might struggle in their first 90 days, and the third handles the entire interview scheduling process automatically.
• Dan: Sounds like you've got one of each type. The job posting tool is generative—it creates content. The new hire prediction is predictive—it forecasts outcomes. And the scheduling tool is agentic—it takes action autonomously.
• Jessica: That actually makes it much clearer. So when I present these to leadership, I should set different expectations for each?
• Dan: Exactly. You'll review and refine what the generative tool produces, act on insights from the predictive tool, and set guardrails for the agentic tool.

Notice how Jessica uses the three-type framework to quickly categorize each tool, and Dan reinforces how the type determines the appropriate way to work with each one. When a vendor describes their product, you can now ask: "Is this primarily generative, predictive, or agentic?" That single question clarifies expectations and helps you assess fit for your specific needs. With these distinctions clear, you're ready to apply your knowledge in an upcoming role-play session where you'll practice identifying AI types and explaining their relevance in realistic HR scenarios.