AI-First Leadership: Cassie Kozyrkov Challenges Us to Rethink Decision-making in the Age of AI

What does it really mean to lead in a world where artificial intelligence (AI) is rapidly transforming industries and decision-making processes? That’s the question Cassie Kozyrkov, founder and CEO of Kozyr, and former chief data scientist at Google, urged attendees to consider during the opening keynote at the 2025 INFORMS Analytics+ Conference. With a blend of candor, expertise and strategic insight, Kozyrkov prompted the audience to evaluate whether they are using AI thoughtfully and responsibly or simply relying on it to produce answers without fully understanding the underlying questions.

Kozyrkov’s central message was clear: Being “AI-first” is about revisiting what was once deemed impossible and daring to reimagine the potential of AI. She encouraged the audience to revisit ideas previously dismissed as impractical and recognize how AI can now make them possible.

“Wouldn't it be great if...” thoughts that were once stored away are now achievable through the power of AI.

Describing AI as “an attic full of wishes,” Kozyrkov emphasized that innovation requires not only technological progress but a shift in mindset. Being AI-first is not about forcing a square peg into a round hole. Instead, it’s about approaching AI with the question: How can AI help me do this? She encouraged the audience to experiment with AI by rethinking old ideas and recognizing AI’s capacity to enhance creativity and problem-solving.

Kozyrkov identified three primary ways to communicate with AI: code, prompts and data. Coding offers precise control over AI systems, but the complexity of data often makes it challenging to provide clear instructions. Prompts, on the other hand, allow for intuitive, natural language interactions, giving users an accessible way to harness AI’s potential. Data – the raw material AI relies on to make sense of the world – represents complexity.

However, Kozyrkov warned that technology alone is not enough. AI systems are only as valuable as the purpose and direction set by human decision-makers. No matter how advanced AI becomes, it will always require human judgment to determine what is meaningful and ethical. As Kozyrkov put it, “The purpose is important, and the purpose comes from people.”

Kozyrkov highlighted the critical role of decision-makers in responsibly guiding AI. Although AI can produce endless answers – particularly with generative systems – deciding which answers are useful and ethical remains a deeply human task. AI can be an incredible brainstorming partner, but it ultimately requires people to provide clarity, purpose and oversight.

She further emphasized the importance of collaboration between technical experts and decision-makers. Although developers and data scientists may build sophisticated AI systems, leaders must define what problems are worth solving and ensure AI is applied ethically.

Kozyrkov reiterated that decision-makers should prioritize their most important tasks and direct AI systems toward meaningful and responsible outcomes.

Acknowledging the increasing complexity of AI-driven decision-making, Kozyrkov highlighted the unique responsibility INFORMS members hold as experts in analytics and decision-making.

AI can generate countless potential answers, but it is ultimately the decision-makers’ responsibility to discern the most appropriate and responsible solutions. Leaders must remain adaptable, open to new possibilities and continuously refining their understanding of what AI can accomplish.

In concluding her talk, Kozyrkov introduced her “Four Pillars of Trust” essential for responsibly guiding AI: competence, integrity, benevolence and charisma. Establishing these pillars, she argued, is critical to building trust in AI systems and ensuring their ethical application. Trust, she noted, is not only about delivering results but also about ensuring transparency and accountability in how those results are achieved.

Kozyrkov’s keynote was more than a presentation; it was a passionate call to action. She reminded the audience that AI is a tool designed to empower decision-makers to achieve smarter decisions for a better world – INFORMS’ guiding principle. Technology can help unlock new possibilities, but it’s human leadership, ethics and creativity that will define the impact AI has on our future. The promise of AI is vast, but it requires courageous leadership and unwavering commitment to use it wisely.