The consensus has been clear across cohorts of my AI leadership trainings this year: Leadership is changing, and it needs to be even more about the people we lead when we add new technologies like AI and agents to our teams. But how, and in which ways?
I recently caught up with Danielle Gifford (Managing Director of AI at PwC Canada). As fellow LinkedIn Top Voices, Global AI Ambassadors, and Adjunct Professors, we had quite a lot to talk about and a lot in common. Whether leaders should put business problems before technology, treat agents as goal-driven systems rather than simple automation, or get their organization ready with process maps, role boundaries, and realistic pilots, these are just a few highlights. Here’s what we talked about…
Focus on Business Problems Before Technology
Start with the problem, not the model. Oftentimes, teams rush to test several AI tools that look exciting, only to discover the underlying process is broken. Before you pick an app or model, map the workflow you want to change. Identify the pain points, the data that feeds the work, and the decision points that need human judgment.
That map does two things. First, it tells you where a single agent or a set of agents could actually save time or reduce errors. Second, it exposes hidden dependencies between systems, approvals, and team knowledge that need to be addressed before AI agents can do their job.
For leaders, this means allocating time and budget to process discovery and not just tool trials. It also means setting success metrics tied to business outcomes, such as cycle time, error rates, and revenue influence.
Start small: choose a high-value, well-defined process to pilot. Use the pilot to learn about integration, user experience, and governance. If the pilot shows measurable benefit, scale with the same disciplined approach: clear objectives, repeatable deployment steps, and a plan for ongoing evaluation.
NEW ONLINE COURSE — Mitigate AI Business Risk
Business leaders are under pressure from their boards and competitors to innovate and boost outcomes using AI. But this can quickly lead to starting AI projects without clearly defined, measurable objectives or exit criteria.
Learn how to implement proven risk mitigation strategies for starting, measuring, and managing AI projects. Along the way, get tips and techniques to optimize resourcing for projects that are more likely to succeed.
Understand Agents as Goal-Driven, Contextual Software
Agents are different from traditional rule-based automation. Where classical automation follows explicit scripts, agents are built to pursue a goal, consult context, and choose actions within defined boundaries. That makes them powerful, but also different in how you design, monitor, and govern them.
Treat an agent like a new kind of team member. What goals do you give it? What information can it access? Which systems may it act on autonomously, and where does it need human sign-off? Those answers determine the architecture and controls you must build. You’ll need logging and traceability so that decisions can be audited, and you’ll need limits on autonomy where mistakes would be costly.
When multiple agents interact, the challenge grows: how do they challenge each other’s assumptions instead of simply agreeing? How do you prevent groupthink among models? Design patterns such as reviewers, validators, and constrained collaborators can help. Agents can also be valuable even if they only handle a portion of a task. A well-scoped agent that prepares options or drafts recommendations can multiply human productivity without taking full control.
Thanks for reading The AI MEMO! Share this thought-provoking post with your network.
Prepare Your People and Processes with Pilots That Teach
Introducing agents changes how people work. It forces you to make role boundaries explicit and to rethink evaluations, training, and team design. In many organizations, key knowledge lives in people’s heads or in informal conventions. Agents force that tacit knowledge to surface. Use that requirement as an opportunity.
Start by documenting tasks at a granular level: who does what, which inputs are needed, and how outcomes are judged. From that base, identify where agents can either augment or take on work. Also, plan for training and give people hands-on time with the tools so they learn what agents do well and where human oversight remains essential. That practical literacy beats abstract lectures every time.
Regulatory signals are coming; expect governance expectations to increase. Put basic guardrails in place from the start: access controls, data handling rules, and evaluation criteria for agent outputs. Run pilots that are designed to teach, not just to win a prize. Use the results to refine role boundaries, policies, and the rollout plan. That way, you turn early experiments into durable capability.
Summary
Leaders can make three practical moves now as AI agents enter the workplace, starting with identifying measurable business problems, treating AI agents as goal-oriented systems that need explicit controls, and preparing people and processes through careful pilots and role design. Taken together, these steps move you from hype to repeatable results.
If you treat this as workforce and process work first, the technology will have a real chance to produce useful outcomes. Start with clarity, protect the parts that require judgment, and teach your teams by doing. That approach will help you move from curiosity to steady, useful adoption.
Equip your team with the knowledge and skills to leverage AI effectively. Book a consultation or workshop to accelerate your company’s AI adoption.
Join my bi-weekly live stream and podcast for leaders and hands-on practitioners. Each episode features a different guest who shares their AI journey and actionable insights. Learn from your peers how you can lead artificial intelligence, generative AI, agentic AI, and automation in business with confidence.
December 16 - Jon Reed (Industry Analyst and Co-Founder of diginomica) and I will wrap up 2025 with our own Agentic AI recap and a 2026 outlook. [More details to follow on my LinkedIn profile…]
Upcoming events
Join me or say hello at these sessions and appearances over the coming weeks:
November04 - Keynote at M-Files & Microsoft event in Cambridge, MA.
November 10-12 - Hands-on workshop and track chair at Generative AI Week in Austin, TX.
November 18 - Masterclass on AI Leadership in corporation with the Employment Agency of Saxony-Anhalt (Germany).
December 11-12 - Panelist at The AI Summit in New York City, NY.
March 09-11 - Attending Gartner Data & Analytics Summit in Orlando, FL.
April22-23 - Keynote at More than MFG Expo in Cincinnati, OH.
Watch the latest episodes or listen to the podcast
