All-in On AI

What would it look like if your company didn’t just use artificial intelligence for occasional experiments, but actually ran on it—every decision, every process, every product infused with intelligent systems? In All-in On AI by Thomas H. Davenport and Nitin Mittal, that’s exactly the question asked. The authors argue that the future belongs to organizations that go beyond isolated AI projects to become entirely AI-fueled learning machines, where the blend of human judgment and intelligent algorithms creates radical transformation. They contend that while most companies talk about AI, very few—less than 1%—have gone all the way. But those that do are already redefining competition, value creation, and the way people and machines work together.

You’ll explore what being AI-fueled means in practice: the components that make organizations truly powered by AI, from data and technology to leadership, culture, and ethics. You’ll discover stories of pioneers such as DBS Bank in Singapore, Ping An in China, and Airbus in Europe—traditional companies that transformed their operations by weaving AI into their very DNA. The book argues that being “AI-first” isn’t just strategy anymore; it’s survival. Whether you’re leading a global enterprise or a small firm, the lessons here show what’s required to thrive in the “Age of With,” where humans and intelligent machines actively collaborate.

From Experimentation to Transformation

Many organizations dip their toes into AI through pilot projects or prototypes, but as Davenport and Mittal emphasize, pilots don’t transform businesses. True economic value—and competitive advantage—comes only when models move into full production. AI-fueled companies deploy hundreds of models that reshape business processes, engage customers, and even create new business models. The authors describe how enterprises like Ping An, with massive datasets across insurance, banking, and health care, use machine learning to drive decisions in real time and learn from every transaction. Such organizations don’t just experiment; they evolve continuously, building what the authors call organizational learning machines—firms that learn from both data and human interactions faster than competitors can copy them.

The Human Side of AI Power

A surprising insight that runs through the book is that technology isn’t the hardest part of becoming AI fueled—people are. Leadership, culture, and skill-building determine success far more than algorithms do. The authors highlight leaders like Piyush Gupta, CEO of DBS Bank, who personally drives data transformation and creates incentives for experimentation. His mantra, “ROI too early kills experimentation,” captures the spirit of fearless exploration that fuels innovation. Likewise, companies like Airbus and Unilever retrain thousands of employees to understand data and AI fundamentals, proving that upskilling an entire workforce creates a stronger, smarter organization. The book insists that being AI-fueled means being human-centered, with ethics, trust, and learning embedded in the system.

Strategy Beyond Efficiency

It’s easy to think of AI as primarily about reducing costs and automating processes, but Davenport and Mittal show that the most advanced organizations go far beyond efficiency. They use AI to reinvent strategy itself. Some companies use AI to create entirely new businesses (Loblaw expanding into digital health), others to transform operations (Kroger using data science to redesign grocery experiences), and still others to influence customer behavior (Anthem guiding healthier lifestyles). The transformation isn’t just technical—it’s strategic and philosophical. By turning data into insight and insight into intelligent action, these companies become more adaptive and more creative.

The Learning Machine Mindset

Ultimately, being AI-fueled means becoming a learning machine—a company capable of evolving through feedback loops of data, decisions, and outcomes. DBS Bank’s chatbots, for example, learn from customer interactions to improve satisfaction and efficiency, while Ping An retrains machine learning models continuously to adapt to new markets. This mindset of scalable learning (a concept echoed by MIT’s John Hagel) allows organizations to turn experimentation into everyday operation. The authors conclude that these pioneers aren’t just using tools; they’re redefining what it means to be a company in an era where machines think and humans learn faster together.

For readers, the invitation is clear: Don’t wait for AI to become “standard.” Whether you lead a firm, manage a team, or shape digital strategy, the time to build a culture, strategy, and system that learn intelligently—with and through AI—is now. Because as Davenport and Mittal put it, in the near future, being AI-fueled won’t be about leadership—it will be about survival.

Before an organization can harness AI’s full potential, it must reshape its human systems—leadership, culture, and learning. Davenport and Mittal underline that successful AI transformations hinge not on algorithms but on people willing to lead, learn, and change. Culture, not code, is what differentiates winners from laggards.

Leadership That Signals Commitment

The book opens this human-centered discussion with DBS Bank’s CEO, Piyush Gupta, a lifelong technologist who transformed a conservative bank into a digital powerhouse. When he arrived in 2009, DBS was jokingly called “Damn Bloody Slow.” Today, it’s a global leader in AI-driven banking. Gupta’s strategy? Encourage experimentation by making it safe to fail. By funding hundreds of small AI projects and showcasing experiments every six months, he turned curiosity into culture. As he says, running early failed experiments acted as “signaling tools”—a public declaration that the organization was serious about transformation, not perfection.

Gupta also led from the front on data transformation. Rather than delegate architecture planning, he personally oversaw the bank’s migration from data warehouses to data lakes, the cleanup of 80 million records, and the creation of cloud-based analytics systems. His leadership mantra, PURE—Purposeful, Unobtrusive, Respectful, and Explainable—guides DBS’s ethical data use. It’s a lesson to every leader: commitment needs visibility and action, not just speeches.

Creating a Learning Culture

Culture shifts in legacy firms rarely happen through mandates. The authors describe how organizations use “data fluency programs” to make AI part of everyone’s job. DBS trained more than 18,000 employees in data skills, creating citizen data scientists; Airbus and Shell partnered with Udacity to teach analytics and machine learning to thousands of engineers and line workers. The message is simple: democratize AI. You don’t need everyone coding neural networks, but you do need an organization where people can interpret and use data intelligently.

Companies like Disney push this further through “evangelytics”—cultural evangelism about analytics and AI, combining storytelling, prototypes, and internal events to promote enthusiasm. Deloitte, meanwhile, established its own AI Academy to systematize talent creation, training both its professionals and clients in how to collaborate with intelligent systems. These examples show that AI literacy—not mere technology—builds readiness for transformation.

Managing Fear and Building Trust

Fear of replacement is an obstacle at every level. Gupta reassured DBS employees that “no one has lost employment at DBS because of AI.” Instead, efficiency gains created room for growth—the bank tripled transactions without adding staff. Similarly, Unilever launched programs where employees design their own career paths and digital reskilling, empowering them rather than threatening them. The authors note that treating employees as co-evolving partners in AI transformation builds trust—and trust accelerates adoption.

In short, if you want your company to become AI-fueled, start with people. Teach managers how to lead experiments, engage employees through storytelling and skill-building, and make clear that machines are partners, not competitors. A well-led culture makes even difficult technological change feel like shared progress rather than imposed disruption.

When most organizations think about artificial intelligence, they start from the bottom—tools, technologies, or isolated use cases. Davenport and Mittal turn that perspective upside down: Strategy must drive AI, not the other way around. They describe how leading companies use AI not just to reduce costs but to redefine strategic direction entirely.

Three Strategic Archetypes

The authors outline three strategic archetypes for AI transformation. First is creating something new: launching innovations that open new markets or products. Loblaw, Canada’s largest grocery retailer, used AI to build digital health ecosystems like the PC Health app—merging pharmacy data, wearables, and loyalty programs to reward healthy actions. Second is transforming operations: redesigning how the business itself runs. Kroger did this by combining machine learning, robotic fulfillment, and personalization algorithms to reinvent grocery management. The third is influencing customer behavior: using AI to guide healthier, safer, or smarter decisions, as seen in Progressive Insurance’s driving score or Anthem’s health nudges.

Platforms and Ecosystems

One of the book’s biggest strategic lessons is that AI and platform business models go hand in hand. Ping An, the Chinese financial giant, exemplifies this approach: it built interconnected ecosystems spanning health care, finance, and smart cities, generating data that feeds continuous learning. Airbus created similar ecosystems through Skywise, an open platform where airlines share maintenance and performance data for predictive insights. These models illustrate how AI enables not just automation but collaborative intelligence across industries.

Strategy isn’t just about deciding where AI fits—it’s about recognizing how AI changes what’s possible. Airbus’s predictive maintenance systems eliminate unscheduled repairs; Morgan Stanley’s Next Best Action platform delivers personalized investment insights via AI recommendation engines. Each strategic archetype pushes AI from support role to creative force.

Integrating Strategy and AI

To make AI strategic, senior leaders must link it directly to core business goals. The authors urge executives to ask: “How can AI help us compete differently?” The 2021 Deloitte survey cited in the book found that firms with enterprise-wide AI strategies were far more likely to outperform competitors. Those that used AI to differentiate themselves didn’t just improve operations—they changed customer experience and expectations. AI becomes a strategic engine when companies continually reexamine business models through an AI lens.

For readers designing strategy today, the takeaway is clear: Treat AI not as a tool but as a central capability that rewires how value is created. Build alignment between top-level vision and data-driven execution. When strategy and AI co-design the business, transformation stops being experimental—and starts being exponential.

AI transformation rests on technology and data—but the authors make clear that “technology follows purpose.” You don’t become AI-fueled by buying software; you do it by aligning technology with intelligent decision-making. Davenport and Mittal detail how advanced organizations construct technical foundations that scale AI across the enterprise.

Building the Toolkit

AI-fueled organizations use the full toolkit—machine learning, natural language processing, computer vision, and rule-based automation. DBS Bank’s anti–money laundering systems combine rules with learning algorithms to prioritize suspicious transactions, cutting investigation time by a third. Companies like 84.51° (Kroger’s data arm) use automated machine learning (AutoML) to speed model creation. Their embedded machine learning framework “8PML” streamlines everything from problem framing to model deployment, allowing non-experts to build production-ready AI faster (“data scientists become multiplier engines”).

Scaling Through Infrastructure

To scale AI, leading firms adopt MLOps—systems that monitor and retrain models automatically. Shell applies MLOps to predictive maintenance across 10,000 pieces of equipment daily, generated from 1.9 trillion rows of sensor data stored in hybrid cloud environments using Azure and Databricks. Engineers are trained through Udacity nanodegrees, proving that scaling requires human as well as technical integration. Unilever faces similar scale challenges, adapting its AI models to local market needs across 100 countries—a reminder that globalization requires custom data solutions.

Managing Data as a Strategic Asset

Data fuels intelligence, so the smartest organizations treat it as infrastructure. Capital One’s cloud migration to AWS made real-time streaming analytics possible, converting decades of batch data into instant insights. Companies like Unilever and DBS built lakehouse architectures combining structured data lakes with traditional relational systems—a “single source of truth.” Shell and Airbus show that data-sharing ecosystems transform industries, enabling collaborative analytics that benefit multiple partners. Deloitte’s data experts call this approach managing the data supply chain—focusing less on collection and more on consumption.

The Role of Emerging Tools

Finally, technology evolves rapidly. The authors highlight the rise of AutoML, AIOps, and high-performance computing as forward-looking enablers. Airbus’s AIOps platform monitors terabytes of daily IT data with machine learning to prevent outages across production plants. Deloitte partnered with Nvidia to launch a center for AI computing using GPU-enabled environments for modeling at scale. These examples underline that companies must designate people—technology scouts—to monitor trends, test innovations, and align them with business value.

Technology and data won’t win by themselves. But when combined with purpose, process design, and people, they become the invisible engine of transformation—the “digital spine” fueling AI’s reach and reliability.

Once the technology foundation is laid, the question becomes: how do you mature AI capabilities—and do so responsibly? Davenport and Mittal present a roadmap for building organizational maturity and ethical integrity. AI-fueled companies, they argue, must balance performance with principles.

Climbing the Capability Ladder

The authors outline five levels: Underachievers (experiments only), Starters (few deployments), Pathseekers (some measurable outcomes), Transformers (multiple production systems), and AI-fueled enterprises (integrated across all functions). Ping An stands as the model Level 5 firm. Founded as an insurer, it transformed into a multi-sector juggernaut spanning finance and health. It has over 4,500 data scientists, an ecosystem serving hundreds of millions, and AI platforms like Ping An Brain for risk,, healthcare diagnostics, and smart cities. This maturity isn’t accidental—it’s designed through leadership investment, data expansion, and dedicated research labs.

Avoiding Pitfalls, Finding Shortcuts

Scotiabank’s story demonstrates that late adopters can catch up fast. Initially cautious, the bank accelerated AI implementation under its Customer Insights, Data, and Analytics unit by uniting data architecture and machine learning under one vision. Through its “blue collar AI” philosophy—focusing on practical operational value over research prestige—Scotiabank deployed 80% of models into production within two years. The lesson: maturity doesn’t require glamour; it requires discipline and integration.

Ethical AI and Responsible Innovation

The authors devote significant attention to ensuring AI’s trustworthiness. They outline principles from Deloitte’s Trustworthy AI Framework—fairness, transparency, accountability, privacy, and reliability—and show how leading companies operationalize them. Unilever’s AI assurance process, for instance, scrutinizes every new AI system for ethical and efficacy risk. A computer vision tool for employee attendance was redesigned to require human oversight after ethical review flagged bias concerns. Ping An established an AI ethics committee to enforce human-centered design, while consortia like EqualAI and the Partnership on AI help organizations share best practices. Automation even supports ethics: tools such as DataRobot or Chatterbox Labs can audit models for fairness and bias.

Ethical AI isn’t a constraint—it’s a catalyst for trust. As the authors note, companies that build governance early move faster later because they don’t spend cycles fixing crises. For you, that means embedding ethical checks as early as feature design. An AI-fueled company earns confidence not only from customers but from its conscience.

To prove that “AI-fueled” isn’t theory, Davenport and Mittal take you industry by industry—from groceries to government—and demonstrate AI in action. These examples show how data-driven creativity redefines every sector.

Consumer and Retail

Walmart uses deep logistics AI for fleet optimization and robotic fulfillment, investing billions to automate warehouses with partners like Symbotic. Loblaw combines loyalty data and health-care ecosystems to personalize nutrition and telemedicine. Kroger and 84.51° deliver billions of weekly recommendations using machine learning and predictive inventory models. These retail leaders prove AI’s ability to balance scale and personalization—what the authors call “mass intimacy.”

Industrial and Energy

Shell exemplifies industrial AI, using predictive maintenance and drone-based inspection to cut downtime from years to days. Seagate applies deep learning to detect wafer defects in chip manufacturing, while Ørsted uses data analytics to optimize renewable wind energy. “Smart factories” like Deloitte’s client partnerships integrate sensor data, robotics, and AI-driven process optimization—proof that industry can learn as fast as tech startups.

Financial and Health Services

In banking, Capital One and DBS represent the pinnacle of integrated AI—from predictive fraud detection to conversational assistants. In health care, Anthem’s AI concierge guides millions toward healthier lifestyles; Ping An’s Good Doctor uses AI to triage conditions for 400 million users. Progressive Insurance’s Snapshot program predicts safe driving behaviors, turning data into life-saving insights. These cases show AI’s power to influence human behavior directly.

Public Sector and Beyond

Governments, too, are joining the movement. Singapore’s AI Singapore initiative funds national-scale AI education and transportation optimization. The U.S. Department of Veterans Affairs uses AI chatbots to assist patients, while NASA and the IRS employ automation to improve service efficiency. Even entertainment evolves: Disney’s Genie planning app uses real-time analytics to optimize park experiences, a microcosm of data meeting delight.

Across industries, these examples confirm the book’s central insight: AI’s impact isn’t limited to one domain—it’s universal. Each company that integrates data, automation, and human creativity becomes part of a global shift where organizations act less like bureaucracies and more like adaptive ecosystems.

How do companies actually get there—from early tinkering to enterprise-wide transformation? The authors map four distinct paths, each showing that no organization is too traditional or too small to go all-in.

Deloitte: From People-Powered to AI-Powered

Deloitte’s journey illustrates how service firms reinvent themselves. Long dependent on human expertise, the firm added a technological dimension. Initiatives like the global Omnia audit platform combine automation, predictive analytics, and machine learning to streamline auditing while enhancing accuracy. Deloitte’s consulting arm launched ReadyAI to offer AI capacity-as-a-service to clients, proving that even a people business can scale intelligence. As Managing Principal Jason Girzadas notes, the goal isn’t efficiency—it’s transformation: “We have to be at the forefront of addressing new challenges with AI.”

Capital One: From Analytics to AI Strategy

Capital One’s evolution from data-driven decisioning to full-scale machine learning showcases the financial sector’s potential. By migrating to Amazon Web Services and creating a Center for Machine Learning (C4ML), the bank built real-time analytics capabilities across every product. CIO Rob Alexander calls this “redefining banking as intelligent software.” Every credit decision, marketing offer, or fraud alert now runs through predictive models. As CEO Rich Fairbank envisioned decades ago, data remains the foundation—but AI makes it interactive, immediate, and adaptive.

CCC Intelligent Solutions: From Data Network to Intelligent Ecosystem

CCC transformed the car insurance industry through AI that reads collision photos and calculates repair estimates in seconds. With billions of labeled images and GPU-powered neural networks, its system elevates accuracy and speed. CEO Githesh Ramamurthy’s bold bet on deep learning turned twenty years of data into customer insight. Today, CCC seamlessly connects insurers, parts suppliers, and repair shops through one cloud ecosystem—proof that midsize firms can become AI leaders by leveraging proprietary data.

Well: The Startup Path

Finally, the health-tech startup Well shows the opposite: what happens when you build AI from scratch. Founded by Gary Loveman, former Caesars CEO, Well uses machine learning and behavioral science to personalize health advice and rewards. Free from legacy constraints, its technologies predict adherence patterns, deliver preventive nudges, and evolve in real time. Loveman’s insight after corporate experience is telling: “Being free of legacy architecture means being free to imagine.”

Together, these journeys reveal essential principles for your own path: know your goal, start with analytics, clean your data, build modular systems, democratize AI skills, and integrate ethics early. Whether you begin with legacy overhaul or greenfield innovation, what matters most is commitment—the decision to make AI not a project but the lifeblood of how your organization learns and leads.

The authors close with both optimism and urgency: AI is no longer optional. Data volume and business complexity are exploding, and intelligent learning is the only sustainable way to manage both. As they write, “In the near future, being AI-fueled won’t be a sign of success—it will be a condition for survival.”