AI Agents: The New Workforce Reality

Forbes

09-05-2025

Founder & CEO of Workmetrics, a leader in workforce software. Doctor of Information Technology specialising in data integration and AI.
They learn, adapt and interact. Operating in a Navy SEAL-like fashion, AI agents have quickly established themselves as one of the most exciting ways to tap into AI for handling everyday tasks. Unlike their assistant siblings that function via scripted response, AI agents understand the context of the environment and make decisions single-handedly to reach specific outcomes—all the while continuously learning.
As it happens, their reactionary nature is having a growing impact on workforce management.
Look no further than the Employee Self-Service Agent, Microsoft's feature within Microsoft 365 Copilot to make HR and IT tasks quicker and easier for employees. It helps them find info and resources faster and overall provides a better experience. For instance, it can pull up the right company policies and resources and combine data from HR and IT so employees have one place to go to for answers.
Speaking of policies, AI agents let employees know when rules or programs change, so everyone stays up to date and follows them. They also automatically keep an eye on whether everyone's following labor laws and company rules so that the company as a whole stays compliant and trouble-free.
From an engagement perspective, AI agents make onboarding smoother and get new employees excited (as much as that's realistically possible) with personalized help right from the get-go.
They not only answer questions contextually but also provide customized advice and help with important stuff like background checks, training suggestions, setting up profiles, and so on. In doing so, they cut down on inherent onboarding problems such as miscommunication or missed steps, while helping new hires get comfortable faster.
IBM already uses (and sells, of course) its IBM watsonx Orchestrate agent in HR. Employees can access it from the company's internal portal, where they're free to ask HR-related questions in natural language and get instant support. But what's really interesting is that through API integrations, the AI agent can actually do HR tasks for employees rather than 'merely' provide assistance.
It can set up accounts, make sure everyone gets the same welcome and schedule training, to name a few things. As a result, HR executives can find more time in their otherwise busy schedules to actually connect with the hires and focus on high-value tasks (not that welcoming new team members isn't, but still).
The agent also tackles talent acquisition by handling interview scheduling and follow-up emails so the hiring process goes smoothly. Furthermore, it automates time-consuming tasks like finding, contacting and engaging candidates on career sites. This frees up recruiters to spend more time on the important aspects—most notably, building relationships.
Besides automated actions and responses, the true power of AI agents lies in predictive capabilities.
Once a potential risk is identified through historical incident reports, IoT sensor data and behavioral analytics, AI agents can use that information to provide instant feedback to the worker. They can do so by dynamically displaying real-time safety information, evacuation routes, and hazard warnings or delivering voice alerts and instructions through speakers or personal headsets.
More impressively, they can adjust the working environment. Say a sensor detects higher-than-normal levels of hazardous gases or excessive heat—an AI agent can increase ventilation and lower the temperature. They can also automatically slow down, pause or completely shut down machinery to prevent accidents if they detect potentially unsafe operating conditions.
Some companies are also integrating AI with digital twins, creating real-time workplace simulations to test safety protocols and identify vulnerabilities before they lead to injuries. AI agents can then use the results of those simulations to create safer environments, all the while aligning with regulatory frameworks such as the EU's Vision Zero initiative and ANSI's Z10 occupational health and safety guidelines.
Granted, AI agents need to be integrated with existing systems to effectively complement the safety team's job. And yes, this doesn't bode well for businesses with legacy infrastructure not quite designed for easy AI integration. But then again, having these intelligent controllers is worth the upfront investment and planning, especially since they can prevent the (previously) unpreventable.
The agents turn everything into organized insights and produce patterns, actionable advice and a basis for human-led decisions driven by easily understandable data. They can also look at lengthy reports, spreadsheets, and legal documents, then pull out the important parts and deliver simple yet meaningful summaries.
That is in the very essence of what AI agents do: break complex workflows into smaller, manageable steps. It means they often specialize in particular areas and tasks. One can dabble in research, another one in multistep customer responses. Plus, you can build a custom agent for a specific need.
To take advantage of what is arguably the most transformative power within the workplace right now, you'll need a solid strategy. Here are some key aspects to handle:
• Analyze your current organizational model: A thorough assessment of your core workforce processes and workflows will help uncover areas where AI will be of utmost value. This includes how and where particular AI agents can be deployed and in the most efficient manner, along with how your company model should accommodate AI-driven demands to deliver optimal results.
• Define tasks suitable for agents: Take a look at your team setup and figure out what will stay with people, what goes to AI, and when you'll still need the old-fashioned brain to step in and take charge. For example, despite their sophistication, AI agents aren't well equipped to deal with tasks involving nuanced human emotions or ethical situations. So, anything involving emotional intelligence or complex social interactions will likely still need a human at the helm to deliver the necessary understanding.
• Support your workers: It's all too easy for your workforce to feel threatened by AI. So, get them ready and comfortable to team up with AI agents by providing them with new skills and ways of doing things. Help them see the roles they can jump into in this AI-powered world so they're less worried about being replaced. Additionally, encourage them to think critically about what the AI spits out so they can jump in and correct it when needed, as well as share their ideas.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?