Managing with AI: The New Role, Skills and Best Practices for Leaders

AI is already reshaping how managers work, decide and lead. The real challenge is no longer whether to use it, but how to integrate it without losing what makes leadership truly effective.

Managing with AI does not mean becoming a technology expert. It means learning how to combine artificial intelligence with human intelligence. The challenge for organizations is not simply selecting the right AI tools, but understanding how they reshape managerial practices, decision making and relationships with teams.

Laurence Ballereaud and Bertrand Déroulède, management experts at Cegos, share practical advice, useful tools and the mindset leaders need to turn AI into a genuine management ally without losing sight of what matters most: people.

The Augmented Manager: A New Role in the Age of AI

An augmented manager is someone who combines the capabilities of artificial intelligence with human intelligence to lead more effectively. Rather than being replaced by technology, they learn to orchestrate it.

In practice, this means evaluating tasks according to two dimensions: their added value and their potential for automation.

This framework helps managers identify quick wins while preserving the activities that generate the greatest value and should remain at the heart of leadership.

What AI Changes for Managers

Leading a Hybrid Workforce

Managers are no longer leading only teams of employees.
They now operate within a hybrid ecosystem where contributions come from team members, AI systems and fully automated tools. Their role increasingly involves coordination, supervision and arbitration. They must define who does what, when and under what conditions.

"The challenge for managers today is not becoming AI experts. It is becoming comfortable enough with AI to combine technological literacy with a genuinely human leadership approach."
Laurence Ballereaud, Director of Management & Leadership Training Projects at Cegos

The more powerful technology becomes, the more managerial value shifts toward what cannot be automated: human capabilities.
Fortunately, AI can significantly reduce administrative workload. The time saved can then be invested in supporting employees, strengthening team cohesion and developing a coaching-oriented leadership style.

Critical Thinking and Discernment: Essential Skills in the Era of Generative AI

AI generates content quickly, but not always accurately.
Managers must systematically assess the reliability of AI-generated outputs, question assumptions, challenge recommendations and explore alternatives.

They must also remain vigilant against their own confirmation biases. People tend to trust information that confirms what they already believe. Generative AI can reinforce this tendency because it often validates rather than challenges the user's assumptions. This risk becomes even greater when algorithms reproduce or amplify existing biases embedded in data.
For example, if you ask an AI tool why a particular employee is underperforming, it is likely to focus on finding explanations rather than questioning whether the employee is actually underperforming in the first place.

In this context, managers are no longer valued because they know more than everyone else. Their value lies in their ability to distinguish reliable information from unreliable information.

“Discernment is becoming the manager's most important skill. It helps combat what I call ‘artificial laziness’.”
Bertrand Déroulède, Head of the Management Training Offer at Cegos

Making AI Your Everyday Management Ally

AI for Production: Optimizing Time and Efficiency

Generative AI is particularly effective when it comes to producing content. It can summarize documents, create presentations, draft meeting minutes and even generate videos. This is currently the most common and widespread use of AI in the workplace, and the productivity gains for managers can be substantial.

However, leaders should carefully select the use cases they delegate to AI and avoid transferring editorial responsibility to the technology. AI can draft, but managers remain accountable for the final output.

AI for Preparation: Expanding Your Thinking

When facing complex situations, limited time or a lack of opportunities to discuss ideas with peers, AI can become a valuable thinking partner.

It can help managers prepare for difficult conversations, deepen analyses and explore multiple scenarios before making decisions. By testing different approaches and arguments, leaders can better anticipate reactions and outcomes.

AI can also provide alternative perspectives, suggest questions that may have been overlooked and help managers step back from immediate concerns to consider broader implications.

Used wisely, AI supports reflection rather than replacing it.

AI for Evaluation: Recognizing Individual Differences

Artificial intelligence can support employee evaluation by continuously analyzing performance data and identifying patterns. however, managers must interpret, qualify and contextualize these insights before drawing conclusions. Managers understand individual circumstances when AI identifies statistical deviations.

Consider a customer service representative whose calls are longer than average and whose conversion rate is lower than that of colleagues. An AI system might classify this as underperformance. In reality, that employee may be handling the most dissatisfied or complex customers, requiring additional time while creating significant value for the organization.

Without human judgment, the data alone can lead to misleading conclusions.

AI for Decision Making: Retaining Human Accountability

AI is highly effective at generating recommendations. Making decisions is another matter.

Managers remain legally and ethically accountable for their choices. Responsibility cannot be delegated to an algorithm.

Moreover, AI lacks emotions, intuition, relational intelligence and lived experience. It does not possess a deep understanding of organizational dynamics or the nuances of a particular context. These human dimensions are often critical when making sound decisions.

Even when AI-generated recommendations appear logical, managers should remain willing to challenge them and trust their own judgment when circumstances require a different course of action.

AI Tools and Use Cases for Managers

For leaders looking to move from theory to practice, a growing ecosystem of AI tools can support a wide range of management activities.

Writing and Documentation

ChatGPT and Claude: summarizing meetings, drafting sensitive emails, preparing performance reviews and one-to-one discussions

Otter.ai and Fireflies: automatic meeting transcription, generation of meeting summaries and action items

Management and Decision Support

Tableau and Power BI with AI capabilities: data visualization, detection of weak signals and emerging trends, automated dashboards

Microsoft Copilot: integration with Teams, Word, Excel and other Microsoft 365 applications for productivity enhancement across daily managerial activities

Learning, Development and Coaching

OLI (The Manager's Companion): an AI assistant dedicated to management topics, available around the clock to help managers navigate challenging situations

AI-powered HR platforms such as Neobrain and Cornerstone: skills mapping, personalized development pathways,, talent development support

This incremental approach allows managers to build confidence, identify risks and create sustainable habits.

Expert Tips for Using AI Effectively

Laurence Ballereaud and Bertrand Déroulède recommend several practices that help managers maximize the benefits of AI while preserving critical thinking and accountability.

Start with Your Own Thinking
Before consulting AI, write down your own analysis or initial ideas.
This ensures that technology supports human thinking rather than replacing it.

Use AI in Areas You Understand
Managers are unlikely to detect errors in domains they do not know well.
The better your expertise, the better your ability to evaluate AI-generated outputs.

Use Indirect Prompting Techniques
Instead of asking AI for answers, ask it for questions.
For example:"I need to present a new organizational structure to my team. What questions are employees likely to ask?" You can go further by asking the AI to present one question at a time and wait for your response before continuing. This conversational approach stimulates reflection and helps leaders prepare more effectively.

Follow a Structured Prompting Method
The quality of AI outputs largely depends on the quality of the prompts.
At Cegos, one recommended framework is the RODEV method: Role, Objective, Details, Example, Verification. This structure helps users formulate requests that are clear, contextualized and verifiable.
Prompt engineering should not be viewed as a complex technical skill. Instead, managers should see it as a practical method for communicating expectations clearly.

Challenge Your Assumptions
One of the most effective ways to reduce confirmation bias is to ask AI to disagree with you.
For example: Ask AI to identify weaknesses in your reasoning. Request counterarguments to your preferred solution. Ask it to organize a debate between two experts with opposing viewpoints.
These techniques encourage broader thinking and reduce the risk of becoming trapped in a single perspective. Used in this way, AI becomes not just a productivity tool but also a catalyst for better decision making.

Supporting Teams Through AI Adoption

Fostering Engagement and Accountability

Beyond new tools and new ways of working, AI is transforming people's relationship with work itself.
“With AI, I am observing a deeper shift that goes beyond usage. It is changing how people perceive work and their contribution.” explains Laurence Ballereaud

Managers play a critical role in helping employees navigate this transformation. Their mission is not only to encourage adoption but also to strengthen engagement, accountability and professional fulfillment in an AI-enabled workplace.

Four leadership levers are particularly important.

• 1. Acknowledge Fears and Resistance
  Like any major transformation, AI adoption naturally generates concerns.
  Employees may worry about job security, loss of expertise, reduced autonomy or declining value in their work.
  Rather than dismissing these fears, managers should encourage open dialogue and create opportunities for teams to express concerns and ask questions. Trust is built when people feel heard.
• 2. Redefine the Value of Work
  AI can dramatically increase productivity. Yet productivity alone is rarely a source of intrinsic motivation.
  Traditionally, professional satisfaction often came from the intellectual process involved in producing a result. Generative AI shortens that process by providing immediate outputs. As a result, organizations must rethink where value is created.
  The focus shifts from producing content to improving it, from generating answers to applying judgment, and from executing tasks to creating meaning.
• 3. Restore Human Agency
  AI provides speed, structure and analytical power. Humans provide meaning, creativity, empathy and contextual understanding. Managers should encourage employees to challenge, enrich and personalize AI-generated outputs rather than passively accepting them.
  When employees become active contributors rather than simple consumers of AI-generated content, they regain a sense of ownership and intellectual engagement.
• 4. Establish Responsible Collective Practices
  AI also raises environmental and ethical questions.
  Training large language models and operating AI infrastructure require significant energy and water resources. Organizations therefore need to encourage responsible usage.
  Managers can help by promoting simple collective guidelines, such as: Avoiding unnecessary prompts and repetitive queries. Using traditional search engines when AI is not required. Respecting data privacy and security rules. Considering the environmental impact of technology choices.
  Responsible AI adoption is not only a technical issue. It is also a leadership issue.

A Three-Step Framework for AI Adoption

Successful AI integration requires more than enthusiasm. It requires structure. The most effective managers guide their teams through a pragmatic three-step approach.

• Step 1: Establish a Governance Framework
  Before experimenting with tools, define clear rules and boundaries. A governance framework should clarify: What employees are allowed to do, which tools are approved, what data can and cannot be shared, when human validation remains mandatory, how confidentiality and compliance requirements are protected.
  This initial framework reduces risk while creating a safe environment for experimentation.
• Step 2: Start with Use Cases, Not Tools
  Many organizations begin by selecting technology and then looking for applications. A more effective approach is the reverse.
  Start with real operational challenges: tasks consuming excessive time, activities frequently postponed because there is not enough capacity, processes creating frustration for employees.
  Once these opportunities have been identified, teams can test AI solutions against clearly defined needs.
  The objective is not to adopt AI for its own sake, but to solve meaningful business problems.
• Step 3: Share Experiences and Learn Collectively
  AI evolves rapidly. Individual experimentation is valuable, but collective learning accelerates progress. Managers can encourage knowledge sharing by creating communities of practice where employees discuss: successful use cases, failed experiments, lessons learned, effective prompting techniques, emerging risks and best practices.
  Some organizations even develop shared prompt libraries to help employees learn from one another and avoid reinventing solutions.
  The more teams learn together, the faster they develop confidence and maturity in their use of AI.

A 30-Day Action Plan for Managers

Managers do not need to become AI experts before getting started.
A gradual and structured approach is often the most effective.

• Week 1: Personal Exploration
  Identify two or three recurring tasks such as: writing meeting summaries, drafting sensitive emails, preparing one-to-one conversations. Use AI to support these activities while retaining control over the final output. Track the time saved and evaluate the quality of the results.
• Week 2: Identify Team Use Cases
  Organize a short team discussion. Ask each team member a simple question: "Which task consumes the most time in your work?" Together, identify the most promising opportunities for AI support and recruit one or two volunteers to experiment.
• Week 3: Experiment and Share
  Review the results collectively then discuss: what worked well, what did not work, unexpected discoveries, emerging concerns.
  With support from AI if appropriate, draft a simple team charter covering: approved uses, prohibited uses, confidentiality requirements, human validation responsibilities.
• Week 4: Adjust Your Leadership Focus
  The most important question is not how much time AI saves. It is what managers do with that time. Schedule dedicated time for activities that create uniquely human value: coaching conversations, development discussions, strategic reflection, team cohesion, meaningful feedback.
  This is where the true value of the augmented manager emerges.

Artificial intelligence is not replacing managers. It is fundamentally reshaping their role and their contribution. By automating routine activities and accelerating information processing, AI creates opportunities for leaders to focus on what matters most: people, meaning, judgment and decision making.

However, successful AI adoption cannot be imposed from above. It must be built progressively through experimentation, learning and responsible leadership.

Tomorrow's most effective managers will not be those who use the most AI. They will be those who know how to combine technological capabilities with human discernment, establish clear collective frameworks and preserve the uniquely human dimensions of leadership. That is what defines the augmented manager. And that is how AI can become a genuine driver of both organizational performance and human development.

This article was originally published on cegos.fr under the title "Manager avec l'IA : bonnes pratiques, rôle et compétences clés"

Did you find this article helpful ?

Expert

Bertrand Déroulède

Learn more