How Agent Roles Will Change With AI

Last week Jensen Huang laid out the vision that the IT department of every company is going to be the HR department of AI agents in the future. This is likely to be the most profound shift in IT we’ve ever seen, because it completely alters the role of the IT department and its responsibility for overall execution of the company. In the past, we went to IT to procure and deploy software that helps enable employees and power workflows across the enterprise. But it was ultimately up to other functions (from HR to the lines of business) to ultimately drive the outcomes and execution of work in the company. AI Agents flips this all. Now, increasingly, in an AI-first enterprise, we can imagine going to the IT department to actually get the work done with AI in the company. With AI Agents, an enterprise can now deploy any amount of “workers” on a task on demand to solve a specific problem in the business. This could be generating leads in sales, writing code and squashing bugs, reviewing contracts or processing invoices. Now, the business will increasingly go to IT to ask for a particular task or set of tasks to get done, and it’s the IT organization’s responsibility for getting those outcomes delivered. This means IT must be insanely close to the business, understanding all the various needs, connecting the dots to major technology trends, and ultimately implementing the right AI architecture to accomplish this. The success or failure of this work now comes down to AI architectures and the AI stack a company leverages; ultimately the decisions IT makes in AI will determine the company’s effectiveness in execution. This changes IT forever.

AI is going to reshape ICs into managers. Soon, engineers will be managing a team of AI agents from day one. These agents will be super knowledgeable, familiar with every part of the codebase, and fluent in every programming language. So ICs will need management skills immediately, not years into their careers. They’ll need to avoid what I call the “triangle of bad management” (ICYMI, I wrote about this: https://lnkd.in/gQAi3fjY). Here are the skills I see ICs needing to develop to manage their AI agents: 1️⃣ Goal-setting Instead of doing the work, ICs will need to get really good at setting clear goals for their AI agents. So many AI tools today are designed to create just what you ask for, without necessarily achieving the objectives you have in mind. ICs will need to develop the skill of precisely framing scope, expectations, and desired goals. 2️⃣ Effective feedback Evaluating output and providing better guidance upfront will be critical. IMO, chain-of-reasoning models are really valuable, and learning how to prompt them will be crucial in this next era—they let people understand AI thought processes instead of just seeing the final result. ICs will also have to develop frameworks to quickly review the massive volume of work created by their AI agents. 3️⃣ Delegating I think this skill will take on a big shift in this agentic era. Knowing when to hand work to AI agents and when to handle it personally will be essential. ICs will need to develop good judgment about effectively dividing work. My take is that entirely new management systems will appear, and maybe even specialized agents that manage other agents. The winning organizations will be those that help their people prep for this transition, turning everyone into effective leaders of both humans AND AI agents.

Are you ready to lead a team that never sleeps, never tires, and learns faster than any trainee? That’s the new reality. AI agents are no longer just tools, they’re becoming true team members. At IBM and other tech giants are already embedding #AI agents into their operations, automating routine tasks and freeing up employees to focus on strategic work. But here’s the catch: leading AI agents requires a new kind of #leadership. Unlike managing people, these agents need clear instructions, well-defined parameters, and ethical oversight. So, how do you integrate AI agents into your team? - Start with high-volume, low-variation processes. Think email triage, data extraction, scheduling, draft generation, and report creation. These are ideal first targets for automation using AI agents. - Deploy AI agents with clear goals. Use purpose-built solutions (e.g., email copilots, customer service bots, data analysis assistants) and train them with real data and business context. Avoid blind trials - set measurable outcomes like time saved, accuracy, or end-user satisfaction. - Upskill your team to work in synergy with AI. Automation isn’t enough — you must redefine human roles. Develop skills in prompting, critical thinking, AI supervision, and refining outputs. Your team’s new role: orchestrating intelligent workflows, not just completing tasks. - Establish a continuous learning and improvement cycle. Track performance, gather team feedback, and refine prompts, data inputs, and integrations regularly. Strategic alignment doesn’t happen on autopilot - it requires constant review and clear governance. Remember: AI isn’t here to replace - it’s here to amplify. The future belongs to #leaders who can fuse cutting-edge technology with human talent. Save this post and share it with other leaders ready to embrace the transformation.

Nearly 60% of leaders say AI agents will significantly reshape their org structures within 3 years. But most companies still lack the infrastructure to manage a workforce that isn’t human. AI agents aren’t just tools, they’re becoming your next workforce. From booking meetings to generating revenue, nonhuman participants are already contributing real economic value. And we’re only at the beginning. Here’s a breakdown of the shift Rex Woodbury outlines in his latest piece: 1. The rise of the Agentic Workforce has begun. AI agents are no longer assistants. They’re doing the work. Human-centric models are giving way to agent-managed workflows that scale exponentially. Companies must rethink how work is defined, measured, and rewarded. 2. SaaS platforms will be forced to evolve. Today’s software assumes a human is behind the keyboard. But what happens when AI agents make the decisions and use the tools? We’ll see a wave of infrastructure disruption, built for agent first, not human first, interactions. 3. Economic value without a W2 is here. AI sales agents are already closing deals. But who gets taxed? Who owns the output? This shift is pushing policy frameworks beyond their limits, and regulators are nowhere near caught up. 4. The new “co-worker” dynamic will take time. What happens when teams are half human, half agent? Decision making moves faster, but blame gets blurrier. Organizations need to prepare for A2A (agent to agent) workflows and A2A mistakes. 5. Infrastructure for agents is the next frontier. The first wave of innovation focused on building the agents. The next wave will be about supporting them: onboarding, identity, metrics, and more. Who’s building the QuickBooks or Workday for the agentic workforce? If you haven’t read Rex’s breakdown yet, now’s the time. This is one of the clearest visions of how work is being reshaped in real time. Thank you Rex Woodbury for helping us see what’s ahead. Check out the full piece in the comments below. How is your team preparing for the rise of the agentic workforce? #AI #FutureOfWork #PeopleAnalytics #AIinHR #HRAnalytics

🧠 If AI agents can do 80% of your job... What exactly is your job title now? That question stayed with me. Because this isn’t just about automation anymore. It’s about identity. Over the last 20 years, I’ve helped companies unlock value with AI. But this moment feels different. AI agents aren’t just helping us work faster — they’re starting to own the work: → Drafting strategies → Leading meetings → Making financial decisions → Even hiring contractors and reallocating budgets And they’re learning — fast. Every prompt. Every project. Every outcome. I’m no stranger to transformation. But this shift is so fundamental, it’s rewriting job descriptions before we even have time to update LinkedIn. 📊 What’s happening now: 80% of knowledge workers already use AI to complete tasks AI agents now execute end-to-end workflows with limited oversight Companies report up to 500% productivity gains Entry-level roles in consulting, finance, and project management are vanishing Titles like Junior Analyst or PMO Coordinator may not survive 2026 In IRREPLACEABLE, we describe this as the human shift. But how we navigate it matters. 📚 And now, we have data to back it up. A groundbreaking new study from Stanford University introduces the WORKBank, surveying: → 1,500 workers → 104 occupations → 844 tasks → Alongside 52 AI experts Here’s what it found: ✅ 46% of workers want AI to take over repetitive, low-value tasks 🟥 But many don’t want AI in areas requiring judgment or human interaction 🟨 Critical mismatches exist between what workers want and what AI can do 🧭 A new Human Agency Scale (HAS) helps define how much control humans want to retain over tasks 📈 The biggest shift? From information skills → interpersonal skills This isn’t just a tech upgrade. It’s a realignment of the core competencies that define our value at work. ✅ To stay ahead, I’m doubling down on: Human-AI collaboration fluency Strategic thinking that AI can’t replicate Ethical oversight and empathy Becoming the bridge between human vision and agent execution 💥 So let me ask you: If an AI agent does 80% of your tasks… What’s your role now? Coach? Strategist? Orchestrator? Or something entirely new? 👇 Let’s debate. How are you preparing? #AI #FutureOfWork #AIagents #WorkplaceTransformation #JobTitles #Automation #IRREPLACEABLE#Stanford #WORKBank #HumanAgency #AIleadership

A new Stanford report just dropped: “Future of Work with AI Agents.” It’s one of the most explicit, data-driven pictures yet of how AI is reshaping work, not in theory, but on the ground. Read the report here: https://lnkd.in/gQFgWhek Key insight? Workers want AI to take the grunt work. 46% of tasks are ripe for automation, not because of hype, but because people are ready to offload the low-value, repetitive stuff. Here is my take: We’re not just automating tasks anymore. With large models gaining advanced reasoning capabilities, we’re starting to augment and, in some cases, replace high-skill, creative, and leadership work. The line between assistant and autonomous actor is blurring fast. This moment calls for a new mindset: - Not AI as a tool, but as a teammate - Not just productivity, but partnership - Not if AI can do the job, but how it should Stanford introduces the Human Agency Scale, a 5-level framework for task-AI collaboration. It’s a timely guide moving from back-office automation to front-line augmentation of knowledge, creativity, and judgment. Executives should pay close attention. This shift will redefine roles, incentives, and what “high-performance” looks like. The upside? Enormous. The challenge? Getting the human-agent chemistry right.

Update: Human-AI Task Scale has expanded. I'm thrilled to share the latest update to the Human-AI Task Scale, now with three critical new agent orchestration levels (Levels 8-10). As AI and machine learning continue to transform Learning & Development (L&D), our roles are evolving rapidly. The newly defined levels represent advanced human and AI collaboration, emphasizing a sophisticated balance between human oversight and AI autonomy. These advanced levels illustrate the pivotal new role emerging in our field—the Human-Machine Performance Analyst™ (HMPA). As we transition towards higher AI autonomy, the HMPA's responsibilities shift from content creation to strategic oversight and ecosystem governance, significantly elevating our impact on organizational performance. Quick Overview: Level 8 (Human-Orchestrated Agent Constellations): HMPAs actively design and coordinate specialized AI agents to enhance immersive learning experiences, such as creating dynamic roleplays and performance assessments. Level 9 (Agent-Agent Collaboration with Human Governance): AI agents independently negotiate and collaborate in real time, guided by the HMPA, who ensures alignment with strategic learning outcomes, relevance, and fairness. Level 10 (Agentic Ecosystems with Distributed Oversight): Autonomous AI ecosystems dynamically self-configure learning environments across departments, with HMPAs setting strategic behavioral policies and governance frameworks, rather than managing individual tasks. As Learning & Development evolves, embracing this strategic role will be crucial. I invite all L&D professionals to explore these advancements and consider how to integrate this powerful human-AI collaboration into your practice. How do you see these advancements reshaping your role or your organization's approach to learning? Special thanks to Kim Denton, MA, CSM, MS, (Cohort #1) for inspiring this post!

Day 3 → How Agentic AI Changes the Future of Work (and Workforce Planning) Agentic AI isn’t just changing what we automate — it’s changing what we delegate. We’re not just speeding up work anymore. We’re handing work over to autonomous systems that can think, act, and coordinate across apps and teams. What does that mean for jobs? 📉 Repetitive work is already disappearing. According to McKinsey, roles involving data collection and processing will shrink by over 30% by 2030. 🧾 Think: report generation, scheduling, status updates, basic research — Previously done by entry-level analysts or coordinators. Now handled by AI agents that run 24/7. 📈 But this isn’t just a loss. There is a rising demand for roles in: AI oversight and prompt design Cross-functional orchestration roles Creative and strategic planning Change management and enablement 👥 So, for leaders, workforce planning needs a mindset shift: Not: “Who do I need to hire next?” But: “What capabilities can my agents own — and how do I reskill my team to lead them?” 🧭 This is the new skill economy: Teaching team members how to design tasks for AI Curating goals, exceptions, and escalation paths Moving from being “in the loop” to being in command Companies that adapt early will scale faster — with leaner teams, less overhead, and more adaptive strategies. #AgenticAI #workforce #future