The AI augmented leader: Why leadership is not replaced, but re-written

Artificial intelligence (AI) is no longer a niche capability. Stanford's 2025 AI Index reports that 78 percent of organizations used AI in 2024, up from 55 percent  the year before. And investment is following adoption: the Index estimates generative AI attracted $33.9 billion globally in private investment in 2024.

This pace creates a new type of leadership mandate with two responsibilities: First, taking an enterprise view to enable AI across the organization, and second, taking a very personal view to use AI and becoming a better leader yourself. As clearly, the goal is not to delegate leadership to algorithms, but to expand your capacity while keeping human judgment and accountability close to your chest.

Let's look at this two-fold leadership mandate in more detail.

Join the PEX Network community

Don't miss any news, updates or insider tips from PEX Network by getting them delivered to your inbox. Sign up to our newsletter and join our community of experts. 

Learn More

The enterprise responsibility: Enable AI as an organizational capability

Let's start with the "why", as many AI programs stall because their story is fuzzy: "We're doing AI" is not a strategy in itself – on the contrary. Leaders must create a narrative that clarifies why they focus on AI, and what they are looking to get out of it. And such stories may vary widely: From advancing quality and speed, reducing outsourcing cost, listen better to customers, being quicker in Operations, identifying fraud cases more thoroughly, whichever it is – state it clearly.

Consequently, this also then determines who, meaning which roles or teams or departments (if you distinguish at all), gets augmented with AI, and what it will change in their daily workflow. Which, by the way, is one of the most prevalent leadership challenges you will be faced with: to enable organizations working together with digital agents and to evolve their very role with a new digital colleague being with them all the time.

In my very own context in Operations in Corporate Commercial Insurance, the story is augmentation: deploying AI agents to help our Operations colleagues amplifying their immense expertise and experience – to essentially enable our core mandate, which is flawless delivery in Business Services, Credit Control or Claims Operations. And that framing is deliberate. It pushes us to redesign our work (decision support, knowledge retrieval, handoffs, become faster/more precise/more robust in what we do every day) rather than chase automation for its own sake. And even more. One of the lessons we learned over the years, is that building AI comes with two perspectives: a technical and a content one. And oftentimes it is Operations who "train the machine" through annotations, refining prompts, clearer tagging in tighter feedback loops ("human in the loop") as it is them who deeply know system flows, process dependencies, market or client specificities, the nature of our policies, contracts, payment process etc.

The role of generative AI

McKinsey reports that 65 percent of organizations are regularly using generative AI in at least one business function. So, the question is no longer "Should we experiment?" but "Can we scale safely and repeatedly?" And yet AI scales all what's already in the system – which may be good and bad. If data is unreliable, governance is unclear, controls are weak, or upskilling is episodic, AI will amplify those gaps, too. So, the mandate for leadership, again, is an enterprise one and it expands to shaping data quality and ownership; governance for access and use; controls for accuracy and auditability, privacy/security (to mitigate legal and compliance risks because data which should not have landed with AI has been falsely uploaded); and a workforce plan that is role-based and long-term. That is also why we are investing in Strategic Workforce Planning to systematically identify the skills of the future and adapt our training and onboarding structures accordingly.

Finally, at enterprise level leaders need a portfolio approach to AI: Go deep on a small number of lighthouse use cases where you are willing to rethink the way of operating, not just accelerate today's steps. And then essentially for an Operations agent, that means embedding AI into real workflows, defining human review points, and measuring end-to-end value (cycle time, quality, risk outcomes, or employee experience). And then go broad by enabling pretty much all teams to use AI for everyday productivity and domain work – within guardrails. While light-touch use includes clearer emails and meeting synthesis it is the advanced use which accelerates the value of AI: contract comparison, recurring reports, or natural-language access to dashboards and curated data products for team leader's decision taking.

The personal responsibility: Use AI to become the leader you intend to be

Before adding tools to your daily routine, also here it starts with having an answer to your very personal leadership intent. In essence, decide what kind of leader you want to be for your team or organization: a motivator for the future, a focus provider for leaders, a decision-enabler for independent teams, a listener to the ground floors, a connector across levels and departments, a storyteller. Whichever it is – naming your intent turns AI from a novelty into a targeted capability. Which may come in two maturity levels.

First, use AI to buy back time for leadership work that compounds: summarizing, comparing documents, preparing Project Steering Bodies, finding content in excessive reading material, providing industry reviews in executive style, and many more. Tasks, for which historically a Business Analyst or an Executive Assistant would have spent hours and days to prepare suddenly become a daily companion at your very own desk or mobile phone.

A Harvard Business Review article on generative AI-assisted leadership writing highlights an experimental study in which using a generative AI tool made professionals more than 50 percent  faster at daily tasks while improving quality. And yet complementary, Stanford University highlights how AI "slop" masquerades productivity and illusions progress. Meaning that AI not only produces quality outcomes but also content, which is blown-up, unprecise, sometimes even untrue – eventually creating hidden cost of management. So, AI does help to take-over tasks that are time consuming and it concurrently demands leaders to challenge its quality in the first place.

Second, use AI to become better at the parts of leadership you cannot outsource: pressure-testing decisions, rehearsing a town hall narrative, role-playing stakeholder objections, negotiating contracts with tough negotiating partners, and many more. Here, AI can take more of a coaching or even mentoring role. It helps you with curating knowledge, it helps you testing arguments or hypotheses on your strategic thinking, it can even help you accelerate your personal growth as a leader (not good on stage? Train it with the machine). In essence, AI clearly has the potential to accelerate creative thinking, challenging ideas, or even amplifying your abilities as a leader. And it therewith creates space for one asset the machine will never take away: spending time with the people you lead, influencing topics with senior management, or any other social and human interaction that makes the majority of a good leader.