Larry Ellison wants to feed all citizen data into one giant AI system

Imagine a world where the government gathers every citizen's medical records, clinical lab results, and full genomic DNA data into a single master database, and lets AI query it at will.

To privacy advocates, that sounds like a dystopian nightmare. But to Oracle cofounder and executive chairman Larry Ellison, it is simply the missing link in modern governance.

Speaking with former UK Prime Minister Tony Blair at the World Governments Summit in Dubai this week, Ellison argued that while public institutions sit on oceans of citizen records, the information is far too fragmented for artificial intelligence to digest.

Ending the 3,000-database nightmare

Ellison pointed out that today's public data infrastructure is an operational mess. You cannot just point an AI model at one source; instead, officials have to chase down records across 3,000 separate silos.

His proposed solution is radical consolidation. He called on governments to unify healthcare histories, diagnostic files, electronic health records, and even national genomic databases into one centralized system tailored for AI consumption.

Why take such a drastic step? Ellison offered two major selling points. First, he claims access to comprehensive genetic and clinical data would allow AI to develop superior therapies, catch diseases earlier, and massively cut healthcare expenses.

Second, centralized data makes fraud detection far easier. Ellison pointed to the spending inefficiencies uncovered by his close friend Elon Musk at the Department of Government Efficiency (DOGE) as clear proof that automated oversight is long overdue.

He also stressed that countries must build out sovereign data centers on their own soil. Just like seaports or airports, hosting data facilities domestically is the only way to satisfy national security and local privacy constraints.

The giant bullseye for hackers

Of course, stacking all of a nation's most sensitive secrets in one place creates a catastrophic single point of failure.

Federal agencies are already relentless targets for sophisticated threat actors. In fact, US intelligence confirmed that Chinese-linked hackers breached the Treasury Department just last December in a major incident.

Ellison acknowledged security concerns, but brushed off current vulnerabilities by arguing that legacy government software is simply too primitive. In his view, full-scale modernization is the only real fix.

Why it matters

We see this proposal less as pure public policy and more as an aggressive commercial vision tailored directly to Oracle's strengths.

For everyday citizens, handing over irreversible biological data like DNA to a centralized state model is an enormous leap of faith. The societal upside in preventative healthcare is undeniable, but the collateral risk of a catastrophic data leak is almost too high to fathom.

For enterprise tech and cloud providers, Ellison's blueprint represents the ultimate pipeline. Building out sovereign AI data centers and migrating thousands of legacy databases into a unified engine is precisely what Oracle sells best.

Pushing AI into public administration makes sense, but turning genomic records into training fuel pushes the envelope too far. Until security standards can actually guarantee immunity against state-backed intrusions, locking all national data into a single vault looks dangerously premature.