What Most GenAI Strategies Get Wrong
It goes without saying that Generative AI is dramatically changing how many organizations operate. It is already influencing how we create content, assess risk, augment workflows and engage with customers. The excitement is understandable, but many businesses are starting in the wrong place.
Too often, conversations about AI begin with capabilities rather than context. A recent EY survey found that 83% of senior business leaders believe their organization’s AI adoption would be faster if they had stronger data infrastructure in place. The real differentiator lies in the groundwork. That starts with a robust, well-governed data strategy.
Quality Over Quantity
There is a persistent belief that large volumes of data will naturally lead to better outcomes. But GenAI is not a volume game. These systems depend on data you can trust, grounded in facts.
High-quality datasets don’t happen by accident. In our case, decades of investment in structured, licensed content (such as what powers the Factiva platform) have been foundational. To date, Factiva includes licensed content from reliable news sources published in more than 200 countries and over 30 languages with specific GenAI licensing rights from nearly 5,000 publishers. This content is continuously tagged, curated and updated by a team of experienced editors.
That human input ensures not just accuracy, but rich context: a key differentiator when aiming for explainable, high-stakes AI output in fields like finance and compliance.
Governance As the Backbone
Many organizations still treat data governance as a back-office function. It is often introduced after AI tools are already in development or use. But by then, it is usually too late.
Governance must be embedded from the start. That includes understanding where data comes from, how it is labeled, who has access, and whether its use is appropriately licensed and compensated. Clear permissions, audit trails, and fair agreements aren’t just legal safeguards; they’re essential to building systems that respect intellectual property and support sustainable AI innovation.
Strong governance reduces risk. It also builds trust, both internally and externally. That trust is built through systems that are transparent, auditable, and designed with purpose.
Infrastructure For Scale and Speed
A strong data strategy is about more than collection and classification. It is also about infrastructure. Organizations need systems that can move data securely and efficiently across departments and products. Those systems must be adaptable and built to support long-term innovation.
Developing a scalable, secure AI infrastructure takes time and cross-functional effort. But when it works, the impact is measurable. Our work on AI-powered translation at Dow Jones Newswires illustrates this. To deliver over 1,000 multilingual articles daily, we developed custom glossaries, tested translations with native speakers and retained editorial oversight throughout. The result is a system that balances speed with accuracy, an outcome that requires careful planning, not just clever automations.
Building this kind of infrastructure takes time. It also requires alignment between product, technology and editorial teams. But the result is a more agile, more resilient business.
People: The Final Layer of Strategy
No matter how advanced the technology, human oversight remains the most reliable quality control.
At Dow Jones, we focus on Authentic Intelligence: the thoughtful combination of machine learning and human judgment. AI can summarize and predict. But only people review, refine, and ultimately decide.
In our Dow Jones Risk business, for example, systems can flag potentially fraudulent transactions. Analysts then provide the context and determine the appropriate action. In the newsroom, AI tools may assist with research or summarization, but journalists remain accountable for final decisions and reporting.
This balance is especially important as GenAI capabilities accelerate. Joanna Stern, The Wall Street Journal’s Senior Personal Technology Columnist and a leading voice in consumer AI reporting, recently created a short film using a suite of GenAI tools. The result was impressive, but far from perfect. “The AI tools got us 90 percent of the way there,” she wrote. “The final 10 percent—where the story actually made sense—took human work.”
This experiment highlights a larger truth: AI alone lacks the instincts, context and editorial judgment that human professionals provide. When applied thoughtfully, human oversight does more than just mitigate risk. It enhances the quality, relevance and reliability of the final product. That final 10% is where judgment lives. And in today’s world, that’s where value lives, too.
Turning AI Into a Leadership Priority
The rise of GenAI is not just a technical shift. It is a strategic one. Leaders must treat it as such.
That means asking the right questions. Is your data accurate and well-organized? Is it governed in a way that meets regulatory expectations? Do you have the infrastructure to scale AI responsibly? Are your teams aligned on goals and risk management?
These are questions for the boardroom and the C-suite, not just the IT department. The companies that ask them now will be better prepared for what comes next.
Don’t Start With the Model, Start With the Strategy
Generative AI offers remarkable potential. But it cannot deliver on that promise without the right foundation.
What we’ve learned is this: data quality, governance and infrastructure aren’t optional. They are the prerequisites to anything worth calling intelligence, artificial or otherwise. Just as important is the role of people in shaping and supervising these systems.
So if you want to lead in AI, start where it matters most: with data that’s clean, structured and compliant. With governance that’s proactive, not reactive. And with teams that understand the tech, without forgetting the human.
In short, the real GenAI advantage does not begin with the model. It begins with the strategy
