The AI Arms Race Nobody Talks About

Most conversations about data focus on privacy. I believe that framing misses the larger issue. The real issue is autonomy. Entrepreneurs, executives and consumers understand that companies collect information: email addresses, search histories, loyalty card activity, location data, browsing behavior and social media engagement. They understand that data has value.

What fewer people ask is a more important question: Value for whom?

The common assumption is that data is valuable because it can be bought, sold, licensed or monetized. That is true, but incomplete. The greatest value of data is not the data itself. It is the ability to understand, predict and influence human behavior.

More than a decade ago, while reviewing data and algorithmic outputs with my team, I had an epiphany that changed the way I viewed the entire industry. The data we were collecting had become extraordinarily precise. We were no longer simply identifying trends or forecasting behavior. We were beginning to anticipate decisions with remarkable accuracy.

Then a troubling thought occurred to me. If we could predict behavior this well, prediction was not the destination. Influence was. The systems we were building were not just learning what people might do next. They were learning how to increase the probability that people would take a particular action. In the wrong hands, they had the potential to make marionettes out of people.

At the time, that realization felt unsettling. Today, it feels urgent.

The technology has become exponentially more sophisticated. The data is richer. The models are smarter. The systems are faster. What once required teams of analysts can now be executed autonomously by algorithms and AI systems operating at global scale.

Prediction becomes influence

Every click, purchase, search, location ping and digital interaction contributes another piece to a profile. Over time, that profile becomes increasingly capable of anticipating what a person may do next.

As these systems become more sophisticated, prediction evolves into something more powerful. The objective is no longer simply to understand behavior. The objective becomes increasing the probability that people will behave in a desired way. This is where the illusion begins.

People believe they are exercising free choice because they are presented with options. But they seldom ask who selected those options, who determined what they would see and who decided what would remain invisible. The path may feel self-directed. The environment surrounding that path has already been shaped.

For business leaders, this distinction matters because the same tools used to improve engagement can also cross ethical lines. A recommendation engine can help customers find what they need. It can also narrow their choices in ways they do not recognize. A personalized campaign can improve relevance. It can also exploit emotional vulnerabilities.

The difference is intent, transparency and restraint.

We have already seen what is possible

Over the past decade, several high-profile examples have revealed the power of data-driven influence in politics.

They demonstrated how large-scale datasets, psychological profiling and targeted messaging could be combined to influence behavior at an individual level. The breakthrough was not mass communication. It was individualized influence delivered at scale.

What made these examples significant was not the specific outcome. It was the demonstration that detailed behavioral data could be used to shape decisions in ways many people never fully understood. What was demonstrated in politics can be applied anywhere.

The same underlying mechanisms now appear throughout commerce, media, entertainment, finance and technology. Every recommendation engine, personalized feed, optimization algorithm and engagement model is attempting to answer the same question: What message will produce the desired outcome?

The business model is built around behavior

The scale of investment behind artificial intelligence shows how valuable that answer has become. Technology companies are investing hundreds of billions of dollars in AI infrastructure. Data centers consume vast amounts of energy. Entire industries are being reorganized around artificial intelligence.

Those investments are not being made because data is interesting. They are being made because human attention is one of the most valuable assets on Earth.

The business model is simple: Engagement drives influence. Influence drives behavior. Behavior drives revenue.

That does not make every use of data unethical. Entrepreneurs need data to understand customers, improve products and build better businesses. But leaders should be honest about the power they are wielding.

If your business uses data to shape behavior, you have a responsibility to ask where persuasion ends and manipulation begins.

The next leadership challenge is digital sovereignty

This issue is already shaping geopolitics. Governments debate digital sovereignty, AI regulation, data localization and information operations because they recognize that information has become a strategic asset. Nations understand that data can shape public opinion, social cohesion, economic behavior and political outcomes.

The same dynamics are also shaping individual lives every day. These systems increasingly influence what people see, believe and act upon, frequently without them recognizing the forces at work.

That raises a difficult question for every entrepreneur building with data: When does prediction become manipulation?

This debate is bigger than privacy. It is about sovereignty.

Sovereignty is the ability to govern oneself. It is the ability to determine one’s own course. Throughout history, societies have fought for political sovereignty, economic sovereignty and national sovereignty because people instinctively understand that freedom requires agency.

The digital age has introduced a new dimension of that fight. Data sovereignty is ultimately human sovereignty.

What entrepreneurs should do now

Business leaders do not need to stop using data or AI. They need to rethink how trust is created. For decades, trust in digital systems has been built on contracts, credentials and audits. A partner receives access. A vendor signs an agreement. An employee receives permissions. If something goes wrong, investigators review logs after the fact. That model was built for a slower world.

AI systems, automated workflows and machine-to-machine interactions now operate at a speed where trust can no longer be treated as a periodic review process. It must become an operational capability.

Entrepreneurs should start by asking, “Can every participant in our ecosystem clearly prove who they are? Are permissions explicitly defined, limited and revocable? And can we prove what actions were taken and why they were allowed?”

These questions apply to customers, employees, vendors, partners and AI agents alike.

The next generation of competitive advantage will not come from collecting more data. It will come from creating systems where trust, accountability and authority can be verified in real time. Organizations that solve that challenge will earn something increasingly valuable in the AI era: confidence.

The question before us is larger than who owns our data. The question is whether we will continue to own ourselves.

Most conversations about data focus on privacy. I believe that framing misses the larger issue. The real issue is autonomy. Entrepreneurs, executives and consumers understand that companies collect information: email addresses, search histories, loyalty card activity, location data, browsing behavior and social media engagement. They understand that data has value.

What fewer people ask is a more important question: Value for whom?

The common assumption is that data is valuable because it can be bought, sold, licensed or monetized. That is true, but incomplete. The greatest value of data is not the data itself. It is the ability to understand, predict and influence human behavior.