If the Machines Can Remember, Why Can’t the System?

PABS, Artificial Intelligence, the Right to Science, and the Test of Conscious Global Leadership

By HH Prince DeAndre, Duke of Hanumanteshwar

Somewhere in the world, the next pandemic may begin with something very small.

A patient becomes sick.

A physician notices something unusual.

A sample enters a laboratory.

A machine begins reading it.

Within hours or days, scientists may know they are looking at something the world has not encountered before.

The machine will measure.

The laboratory will document.

The sequence will be generated.

A database may assign an identifier.

Computers somewhere else may begin comparing that sequence against millions of others.

Researchers may begin modeling its structure.

Artificial intelligence may help identify patterns.

Pharmaceutical scientists may begin considering vaccines, diagnostics or treatments.

The science can start moving almost immediately.

And yet one of the most difficult questions now facing governments at the World Health Organization sounds remarkably simple beside all of that technological sophistication:

Can humanity remember who contributed what, who subsequently used it, and whether the benefits that followed were shared?

That question sits near the center of negotiations over the WHO’s proposed Pathogen Access and Benefit-Sharing system, known as PABS.

As I write this, governments are meeting in Geneva from September 14 through September 18, 2026, for the eighth round of negotiations over the PABS Annex to the WHO Pandemic Agreement. WHO describes the purpose of the system as achieving safe, transparent and accountable pathogen access and benefit sharing “on equal footing.”

Those words sound straightforward.

What they require from global leadership is not.

The Question AHF Is Asking

AIDS Healthcare Foundation recently reduced the dispute to one question:

The world shares pathogens. Why aren’t the benefits shared too?

Behind that question is a much larger proposal.

AHF wants a PABS system in which commercial participants cannot simply gain access to globally shared pathogen information and later decide whether or how they will contribute back. It is calling for a binding and enforceable system, mandatory benefit sharing for commercial users, user registration and traceability, technology and know-how transfer, nonexclusive licensing opportunities for manufacturers in developing countries, financial contributions and meaningful accountability.

AHF has also made an important distinction that deserves more attention.

Equity cannot simply mean deciding what percentage of vaccines wealthy manufacturers will donate after a crisis has already begun.

The deeper question is whether more regions of the world will eventually possess the scientific and manufacturing capacity to produce those vaccines themselves. AHF has argued specifically for technology transfer, regional manufacturing capacity and the technical and financial resources required to make that capacity sustainable.

That changes the conversation.

We are no longer talking only about sharing medicine.

We are talking about sharing the capacity to make medicine.

COVID Was the Lesson. PABS Is One of the Tests.

This discussion did not begin in 2026.

COVID-19 demonstrated what extraordinary global scientific collaboration can accomplish.

Scientists identified the virus.

Its genome was sequenced.

Researchers around the world began studying it.

Vaccines were developed at a speed that would once have seemed impossible.

Science worked remarkably fast.

Equity did not move at the same speed.

Manufacturing capacity was concentrated.

Wealthier governments secured early supplies through purchasing power.

Many lower-income countries waited.

The world learned that a scientific breakthrough and equitable access to that breakthrough are not the same achievement.

That history is personal for me as well.

My husband, HRH Prince Manvendra Singh Gohil, has been part of this public-health conversation for years through his work with AHF.

In 2021, during AHF’s Vaccinate Our World campaign, he publicly supported efforts to secure COVID-19 vaccine patent waivers “for the benefit of the global population.” In 2024, he was a named participant in AHF’s international Save Our Society intervention concerning the WHO Pandemic Agreement negotiations and the need for an agreement that worked for lower-income countries as well as wealthy ones.

So when I look at the PABS debate today, I do not see a new argument.

I see the next chapter of an argument the world was already having while COVID was still teaching us what happens when scientific knowledge moves globally but scientific power does not.

The Right to Science Is About More Than Reading the Science

There is also a human-rights dimension that deserves greater attention.

Article 27 of the Universal Declaration of Human Rights says that everyone has the right to share in scientific advancement and its benefits.

That sentence contains two ideas.

Scientific advancement.

And its benefits.

Those are connected.

The United Nations Committee on Economic, Social and Cultural Rights later explained that the benefits of scientific progress include practical applications such as vaccinations. It also emphasized that countries participating in international organizations do not somehow leave their human-rights responsibilities outside the negotiating room. International agreements should help developing countries build their ability to generate scientific knowledge and benefit from its applications.

That changes how I look at PABS.

Imagine a country contributes the pathogen.

Its scientists sequence it.

Its laboratories provide surveillance.

Its population may be experiencing the outbreak.

Its public institutions provide epidemiological information.

The world asks that country to share because global health depends upon it.

Then science succeeds.

A vaccine is created.

A diagnostic is created.

A treatment is created.

At what point does the country that participated in the scientific beginning become merely a customer of the scientific result?

That is not simply an economic question.

It is a question about participation in science itself.

What Would AHF’s Ask Actually Look Like?

Imagine that the next dangerous pathogen is discovered tomorrow.

A laboratory sequences it.

The sequence enters the PABS system.

A strong system would not require a scientist to stop working and spend days negotiating administrative paperwork before the data could move.

The data could move immediately.

At the same time, the digital system could create a record.

This sequence entered the system at this time.

It came from this laboratory.

It came from this country.

It carries this persistent identifier.

Its scientific provenance remains connected to it.

Research begins.

A university laboratory examines the sequence.

A public-health agency compares it against other outbreaks.

Another research team builds an epidemiological model.

A pharmaceutical company eventually uses PABS materials or information while developing a vaccine.

At that point commercial responsibilities can become increasingly important.

The system could associate the relevant commercial actor with binding PABS obligations.

If a product enters clinical development, another record is created.

If it moves toward regulatory authorization, another record exists.

If a license is granted to another manufacturer, that relationship can be documented.

If technology is transferred, that can be recorded.

If a manufacturer owes products, money or another agreed benefit to the PABS system, fulfillment can be recorded too.

Science keeps moving.

Accountability moves with it.

That is very different from expecting scientists to manually document every movement of knowledge themselves.

The Digital Systems Already Know How to Remember

This is where I believe some of the PABS debate risks becoming unnecessarily trapped in an old way of thinking.

Modern digital systems already use persistent identifiers, timestamps, metadata and provenance.

The World Wide Web Consortium created a provenance model specifically for representing relationships among data, activities, responsible agents, derivations and time.

The Global Alliance for Genomics and Health has gone even further in its responsible genomic-data framework. It calls for accountability mechanisms capable of tracking chains of data access or exchange to their source and recognizes the importance of maintaining attribution in downstream uses where appropriate.

We should be precise here.

That does not mean every existing pathogen database currently records the verified identity of every person who downloads every sequence.

Many scientific databases were intentionally designed around open access.

Nor does it mean we can perfectly reconstruct every scientific influence behind a product developed from thousands of datasets, publications and experiments.

We cannot.

But those realities do not establish that provenance is impossible.

They establish that governments must decide what needs to be recorded, at which point, and for what legitimate purpose.

That is governance.

Complexity Should Trigger Better Design, Not Institutional Amnesia

I keep returning to something that seems almost obvious.

Humanity already builds extremely complicated systems when institutions decide the information matters.

Banks record financial transactions across continents.

Shipping companies track packages through international supply chains.

Airlines record passenger movements and reservations.

Pharmaceutical manufacturers already maintain extensive production records because quality, safety and regulation require them.

Companies maintain international intellectual-property portfolios.

Research repositories assign persistent identifiers to scientific work.

Digital platforms can process billions of events.

None of these systems is effortless.

None is perfect.

We built them because there was sufficient reason to build them.

So when pathogen information capable of contributing to lifesaving products and enormous commercial value enters the discussion, I think it is reasonable to question the assumption that preserving meaningful provenance is somehow beyond our technological imagination.

The question should not be:

Can we track absolutely everything forever?

Of course we cannot.

The better question is:

Which moments matter enough that humanity should refuse to forget them?

Where did the scientific contribution originate?

When did it enter the system?

Which identifier belongs to it?

Which commercial users entered obligations connected to the system?

Were those obligations fulfilled?

That is a very different standard.

This Is Where AI Can Help, But AI Should Not Be the Judge

Artificial intelligence introduces another possibility, but it also introduces a temptation we should resist.

AI should not become a machine that declares:

Company A owes Country B because the algorithm says Product C came from Sequence D.

That would give a computational model legal and moral authority it should not possess.

The underlying evidence should remain stronger than that.

Authenticated records.

Identifiers.

Timestamps.

Scientific citations.

Contracts.

Regulatory filings.

Licensing records.

Human decisions.

AI can then help humans work across that record.

It can identify relationships among enormous quantities of information.

It can flag a missing identifier.

It can identify where a provenance chain appears to break.

It can notice that a publication, patent, dataset or commercial development appears connected to earlier PABS materials.

It can help auditors prioritize what deserves human review.

It can identify inconsistencies.

It can make a complicated record understandable.

AI becomes a learning and analytical tool.

Machine memory. Human judgment.

That distinction matters.

Open Science Does Not Require Forgetfulness

Industry has legitimate concerns that also deserve to be represented fairly.

The International Federation of Pharmaceutical Manufacturers and Associations argues that free and unhindered access to pathogens and sequence information is essential to fast research. It warns that overlapping legal obligations and complicated access requirements could slow innovation, discourage participation and interfere with the scientific collaboration the world needs during an emergency. It favors open exchange and more flexible, collaborative arrangements.

That concern should not be mocked.

During the opening hours of an outbreak, scientists should not be waiting while lawyers negotiate terms.

But I do not accept that the only alternatives are unrestricted scientific access with weak accountability or strong accountability that paralyzes science.

A mature global institution should be able to design beyond that binary.

The scientific work and the accountability record should be capable of moving at the same time.

Open science does not require institutional amnesia.

A publication can be freely available and still identify its authors.

Open-source software can be freely used and still preserve a record of contributors.

Scientific data can carry persistent identifiers.

Knowledge can move without becoming historically anonymous.

If science can remain open while preserving attribution in so many other circumstances, PABS negotiators should at least ask how much of that architecture can be adapted here.

Traceability Must Not Become Surveillance

There is another side to this that conscious leadership must also recognize.

Just because technology can collect information does not mean it should collect everything.

A responsible PABS architecture should not become a global surveillance system for scientists.

The system should record what is reasonably necessary to preserve scientific provenance, accountability and legitimate benefit-sharing obligations.

Nothing more.

Researchers conducting ordinary noncommercial science may require different treatment from corporations developing commercial products.

Data minimization matters.

Cybersecurity matters.

Privacy matters.

Independent oversight matters.

Clear limits on use matter.

An accountability system earns trust only when the system itself is accountable.

That is another reason AI should assist human governance, not replace it.

Technology Transfer Is Hard. That Does Not Make It Optional.

AHF’s larger proposal raises another legitimate challenge.

Technology transfer is not accomplished by emailing a patent.

A vaccine manufacturing process may depend upon specialized equipment, trained personnel, regulatory expertise, quality-control systems, supply chains, tacit manufacturing knowledge and significant capital investment.

Building that capacity takes time.

It costs money.

It requires qualified partners.

It requires sustained preparation between emergencies.

Those difficulties are real.

But there is a strange tendency in global policy discussions to treat difficulty as an argument against beginning.

If the world waits until a pandemic begins to decide that Africa, Asia or Latin America should have greater manufacturing capacity, we have already waited too long.

Preparedness means building before need becomes panic.

A factory, laboratory or trained workforce developed through one pandemic may become part of humanity’s defense against the next.

That is why the benefit of technology transfer cannot be measured only by what happens during one emergency.

The benefit accumulates.

Knowledge stays.

Workers remain trained.

Regulators retain expertise.

Manufacturing networks continue to exist.

The world becomes less dependent on a handful of geographic locations.

That benefits humanity as a whole.

Follow the Leverage

There is also a power question beneath all of this.

Before a country shares a newly discovered pathogen, it has something the world urgently needs.

Once the sample and sequence information move globally, much of that leverage may disappear.

Later, when a vaccine exists, the leverage may sit somewhere entirely different.

With the company holding the product.

With the owner of the manufacturing process.

With governments possessing purchasing power.

With countries containing the production facilities.

A strong PABS system tries to connect those two moments.

It says that access and benefit are not unrelated episodes.

They are parts of one global relationship.

That may be one reason this negotiation is so difficult.

It is not only deciding what gets shared.

It is deciding when power changes hands and what obligations travel with it.

Conscious Leadership Is Not Agreement

I have been thinking about these negotiations through the lens of conscious global leadership.

I do not use that phrase to declare one country conscious and another country unconscious.

Leadership is more serious than that.

Governments have national interests.

Companies have commercial interests.

Scientists have legitimate concerns about research freedom.

Developing countries have legitimate concerns about extraction and exclusion.

International institutions have operational limitations.

None of those interests disappears because we invoke humanity.

Conscious leadership begins when leaders can see all of those realities at the same time.

It asks whether protecting one interest today creates a larger vulnerability tomorrow.

It asks whether the system rewards contribution fairly.

It asks whether power is accumulating somewhere that makes the entire global system less resilient.

It asks whether the institution has actually learned from the last crisis.

That is the standard I think PABS should be judged against.

Not whether one negotiating bloc wins.

Not whether industry loses.

Not whether intellectual property disappears.

Not whether every pathogen sequence becomes restricted.

The measure is whether humanity emerges with a system that allows science to move rapidly and makes reciprocity real.

Can Global Institutions Learn?

WHO Director-General Dr. Tedros Adhanom Ghebreyesus said this week that cooperation is essential to pandemic preparedness and called for completion of a safe and effective PABS system. He urged governments to build resilience before the next emergency, direct innovation toward places markets miss and make international cooperation operational rather than merely rhetorical.

That last idea deserves attention.

Make cooperation operational.

We use the language of cooperation constantly.

Solidarity.

Equity.

Partnership.

Shared responsibility.

But institutions reveal their values through architecture.

A conscious institution does not merely remember the last crisis.

It changes the architecture that allowed the injustice to happen.

COVID showed us that scientific speed without equitable distribution leaves millions vulnerable.

PABS gives governments an opportunity to decide what they learned from that experience.

Did we learn only that vaccines need to be developed faster?

Or did we learn that access, manufacturing, technology, attribution, trust and global participation are also parts of pandemic preparedness?

What Are We Not Talking About?

This is the question I always return to.

What is sitting in plain sight while attention is directed somewhere else?

In this case, I think we may be spending too much time asking whether traceability is technologically difficult and not enough time asking why technological difficulty is treated differently depending upon whose interests are being protected.

We tolerate enormous complexity to protect ownership.

We tolerate enormous complexity to move capital.

We tolerate enormous complexity to manage international commerce.

We tolerate enormous complexity to document pharmaceutical quality.

We tolerate enormous complexity to enforce intellectual property.

When the discussion turns to documenting contribution and reciprocity, complexity suddenly becomes a reason for hesitation.

That does not automatically make the objections dishonest.

But it does make the contradiction worth examining.

Because if the international system can build sophisticated structures to remember who owns scientific value, it should be capable of seriously exploring sophisticated structures that remember who helped create it.

The Real Measure of PABS

In the end, I do not think the real measure of PABS will be found only in the percentage of vaccines that WHO can distribute during the next pandemic.

That will matter.

Lives will depend upon it.

But there is another measure.

When the emergency ends, how many more places on Earth will possess the ability to respond?

How many more laboratories will be capable?

How many more scientists will have participated?

How many more manufacturers will understand the technology?

How many more countries will possess regulatory expertise?

How many more regions will be able to produce medical tools rather than wait for them?

How much more willing will governments be to share the next pathogen quickly because they trust the system that follows?

That is resilience.

And that may ultimately be the most important benefit humanity can share.

Not only the vaccine.

Not only the treatment.

Not only the diagnostic.

But the knowledge, infrastructure and scientific power to create them.

The machines can record.

Digital systems can remember.

Artificial intelligence can help humans understand the relationships among what has been recorded.

Technology is not perfect, and it cannot answer our moral questions for us.

That responsibility still belongs to people.

The question confronting the WHO negotiations is therefore larger than whether the technology exists.

It is whether global institutions are willing to use the tools already available, improve the tools that are not yet good enough, protect open science, respect privacy, sustain innovation and still insist that contribution and benefit remain connected.

The next pandemic will not wait for perfect institutions.

It will not wait for perfect technology.

It will not wait for governments to agree on every philosophical question.

We must build for the imperfect world we actually inhabit.

Let the scientist work.

Let the information move.

Let the digital system preserve the record.

Let AI help us understand it.

Let commercial responsibilities be clear before an emergency begins.

Let legitimate innovation continue.

Let technology and capacity spread.

Let developing countries participate not only as sources of pathogens and data, but as scientists, manufacturers and creators of the solutions that follow.

And let international institutions prove that they are capable of learning from the crisis we have already survived.

Because the most uncomfortable question facing PABS may no longer be whether humanity possesses the technology to remember.

Increasingly, we do.

The question is whether we possess the leadership to ensure that the system chooses not to forget.