So hi, everyone. Welcome to my session. I want to start with a story. So a few months ago, I was in a meeting.The kind you all, been in at some point. So beautiful dashboards on screen, an impressive, like, live demo, a real AI, system working, miserable, deployed. People are proud, and they should have been. But, six months later, the project had quietly disappeared. No adoption, no ownership, no business impact. Time and money out the window, teams are unhappy, or simply don't care. Does this picture, sound familiar to you? And here is what tells me about that story. It wasn't a failure of technology. The algorithm worked perfectly. It was a failure of leadership. Now I've seen this happen once. I've seen this happen a dozen times, and the data confirms that more than 80% of AI projects fail, and 95% of generative AI pilots never scale.

Not because machines break down, because organizations break down. That is the conversation I want to have with you today. So, I am Charlene Chen. I lead the data intelligence projects at the French Ministry of Economy and Finance. I'm also a member of several expertize in AI networks. I am also an executive MBA candidate at ASEC Business School, where I currently serve as president of the digital and tech club of ASEC Alumni. So my career, has taken me across private companies, public institutions, international environments between supply chains operations and data transformations programs. And what I have learned across all these worlds is simple. Technology alone never transforms an organization. People do. Leaders do. Country does. So today, I want to challenge something that many organizations still deeply believe, something that cuts them millions and quietly destroys their future. AI is no longer a tech topic.

It is a governance issue, a trust issue, a control issue, a business one, and above all, it is a leadership issue. And, like every issue that is truly strategic, it cannot remain only in the hands of the CTO. It belongs to the c suite the CEO's method. Let me ask you something. How many companies today claim to have an AI strategy? Almost all of them. How many are actually transforming? Far fewer. Why? Because they treat AI as a technology deployment problem when in reality it is an organizational transformation challenge. And the research is clear on this. 73% of failed projects lacked executive alignment from the very start. 68% under invested in governance and data foundations. 61% still treat AI as an IT project, not a business transformation, and more than half lost active executive sponsorship within six months. Six months.

Here is what that looks like in the real life. There is a launch. There are workshops, maybe even a LinkedIn post. You know what they want. Then priority shifts. Leadership attention moves on, teams become siloed, each department starts protecting its old perimeter, nobody truly owns the transformation anymore. And the project doesn't crash, it doesn't explode, it just fades. The technology worked. The leadership model around it around it. But is failure the only risk? Now I want to be direct with you. The biggest risk today is not moving too fast with AI. The biggest risk is passive leadership. AI will not replace leaders. But leaders who refuse to own AI will absolutely be replaced by those who do. And this risk is already visible at three levels. The first risk is regulatory. So the EU AI Act is not coming. It's here. It demands traceability, transparency, governance, and accountability. This is no longer a compliance checkbox for the legal stem.

It is an executive responsibility. In five years, boards will ask who signed off on the system and make sure you have the answer. The second risk is, reputational. Imaging this is already here. An AI system produce biased recommendations or leaks sensitive data or automates a decision that harms a customer or an employee, a community, who face the media, not the machine learning engineer. The CEO does because trust has become a leadership currency. And we all know it takes years to build and seconds to destroy. The third risk is strategic. Some organizations still treat AI as a side innovation project. Others are already using it to redesign their entire operations, their customer experience, their decision making, their business models. And the gap between these two groups is growing fast. And at a certain point, it won't become very hard to catch up.

So what does c suite ownership of AI actually look like in practice? I'm not talking about becoming a data scientist. I'm not talking about understanding algorithms. I'm talking about really the seed suit ownership of AI. I'm talking about three things and only three. First, define your principles before a crisis force you to do. What does success look like in your organization? What risk are you willing to accept? Where are the ethical limits? Because if, leadership doesn't define those, guardrails, the organization will improvise them under pressure, and that that is always when crises happen. The second one, break the silence. AI touch, legal, HR, IT, communication, operations, all at the same time. If each department works independently, AI becomes fragmented and dangerous. But when leadership aligns the whole organization around the shared vision, AI becomes a strategic capability. And third, create real account accountability.

One of the deepest problem in AI transformation is what I call the diffusion of responsibility. Everyone contributes, nobody owns, and when nobody owns, trust disappears. Organizations need leaders who are named, KPIs that are measured, and a culture where responsible AI behavior is visible from the very top. Because control always scales faster than policies. Now, I want to address something important, Something that is usually framed as a social justice issue. I want to reframe it today as a performance issue. As you know, we have a paradox of our time. So according to Russel and Reynolds Associates, women hold fewer than 10% of CEO positions in the S and P one hundred and FTS one hundred, precisely at the moment when AI is redefining what leadership means. The 2025 Fortune Global 500 list includes a record 33 women CEOs, representing only 6.6% of the total.

However, studies show that the companies with at least 30% female leadership are 12 times more likely to rank among the top 20% for fine financial performance, according to the Grand Thornton. This correlation becomes really critical in the AI area, where diverse leadership teams help mitigate algorithmic bias, improve decision making, and strengthen long term resilience. And the leadership qualities must most often, associate, associated with women executives as precisely the ones the AI era now demands. Because AI is just not about optimization, AI amplifies system. And when system become more complex, homogeneous leadership become dangerous. The AI Now Institute documented this clearly. The demographic composition of the teams who build and govern AI directly influence algorithmic behavior. AI reflects the people who shape it. So an AI system built without diversity does not come neutral. It simply scales unconscious blind spots at industrial speed. So diversity in AI governance is not cosmetic.

It is strategic resilience. It is risk management. It is competitive advantage. And frankly, we cannot afford to leave that on the table. So now I want to give you something practical before I close. Not a framework, not a methodology, only four questions. Four questions every leader in this room should ask tomorrow morning. First, who truly owns AI governance in my organization? If the honest answer is the IT team or the data department, that is already your warning signal. Second one, have we defined AI principles aligned with our values? Something as concrete as we will not use AI to make HR decisions without human validation. If you don't have that, your decisions will always be reactive, and reactive is expensive. Third, how do we measure AI value beyond ROI, return on investment?

Because AI success is not only cost reduction and automate automation. It is trust, resilience, compliance, customer confidence, reputation. Are you measuring any of that? And fourth and last, do we include diversity in AI design and governance? Not as a diversity checkbox, as a risk reduction strategy. Because diversity is not just representation, it is anticipation. It reduces blind spots before they become crisis. I want to leave you with one image. AI is a mirror, not a magic wand, not a thread from science fiction, a mirror. It amplifies what already exists inside your organization. If you lead with silos, AI will scale those silos. If you lead with short term thinking, AI will accelerate that thinking. If you lead without accountability, AI will scale irresponsibility faster than any technology we have ever seen.

But if you lead with vision, with governance, with trust, if you build teams that reflect the full spectrum of human experience, if you ask the hard questions before the crisis forced you to, then AI becomes something extraordinary. One of the most powerful tools for sustainable value creation, we have ever had access to. So the question is not, who will manage AI? The real question is, what kind of leaders do we want to become in the age of AI? And that question belongs to every one of you, not to your IT team, not to your data scientists, to you. Thank you for your attention, and I will be delighted to discuss with you further.