Why the Future Depends on the Data We Protect

The future of AI won’t be built on more internet data — it will be built on the decades of private, institutional knowledge organizations have been storing offline.

Tape Ark CEO & Founder Guy Holmes was privileged to take the stage at the 2025 RIMPA Global Conference in Melbourne recently, where he shared a powerful perspective on the ethical frontier of artificial intelligence and the critical role of the data custodians who safeguard the world’s most valuable information.

Artificial intelligence may feel omnipresent, but the truth is stark: most of the world’s data has never been seen by an AI model. The documents and emails inside enterprises, medical imagery sitting on private networks, decades of broadcast archives, and the vast government records we preserve every day all exist outside the datasets used to train today’s most advanced systems.

Large language models such as ChatGPT and Gemini are built almost exclusively on public web data — a thin, noisy slice of humanity. The real value, the real intelligence, and the real breakthroughs lie in the private, institutional memory organizations have accumulated for decades.

And that’s where ethical responsibility meets technological opportunity.

AI’s Next Great Leap Depends on Private Archives

Tape Ark’s work around the world shows just how transformative previously inaccessible data can be. In the UK, the team is helping liberate 600 petabytes of historical weather measurements — every observation captured across the northern hemisphere — and making it accessible to modern supercomputing. For the first time, AI and scientific models can analyze this information in full, unbroken continuity.

This is the new frontier. Not scraping more of the internet — but responsibly connecting the vast offline archives organizations have been storing for 50+ years.

Why Governance and Structure Matter More Than Ever

As AI systems begin interacting with enterprise data, records management shifts from a back-office function to a front-line ethical discipline.

AI doesn’t understand meaning; it understands patterns.

It learns not what is true, but what is common.

That means:

• Clean, structured data becomes the “truth” AI sees.
• Messy or biased records introduce risk, distortion, and inequity.
• Metadata, retention policies, and file structures now shape machine behavior.

In essence, records professionals have become the quiet trainers of AI, determining what machines learn and how fairly they learn it.

The Appetite for Data Is Growing — and So Is the Responsibility

Public data is running out. It has been predicted that by 2026, AI models will have consumed nearly everything openly available online. Their next source of intelligence will be private organizational data — but unlocking that data must be done with consent, context, governance, and ethical oversight.

The challenge is not whether AI can access enterprise archives, but whether it should.

From lending decisions to healthcare diagnostics, AI outcomes can either reflect our best values or amplify decades of human bias. The difference lies in how we steward the data.

Why Tape Ark Is at the Centre of This Shift

Tape Ark exists to bring previously unreachable data into secure, modern environments where it can be used responsibly for innovation, analytics, and AI.

By liberating archives from offline media — often tapes that have never been accessible at scale — organizations can:

• Discover insights hidden for decades
• Enable more ethical, accurate and contextual AI models
• Build future technologies on complete, representative datasets
• Maintain strict control over what is shared, what is protected, and what stays private

This is where ethical AI begins: with the custodians of data, and with the infrastructure that preserves its integrity.

The Future of AI Needs Us All

AI is accelerating, but it cannot progress without the stewards of the world’s memory — the archivists, information managers, data specialists, and governance professionals who ensure data is structured, authentic, and protected.

As Guy Holmes notes, “Progress without preservation is amnesia.”
The organizations that thrive will be the ones that bridge yesterday’s knowledge with tomorrow’s technologies responsibly.

Tape Ark is proud to support this global shift — unlocking the world’s most valuable data so AI can learn not just from what is loudest, but from what is true.

 

Watch Guy’s full 20 minute presentation here.

 

Generative AI was used to create this blog post image.