CGI Update on The State of Tech Safety

Two years after launching The Parents’ Network at the Clinton Global Initiative in New York, Prince Harry, Duke of Sussex returned to the stage to update the global community on the state of tech safety. While acknowledging recent legal settlements and regulatory shifts surrounding social media addiction, The Duke warned that technological developments have created an even tougher “uphill battle”. Describing AI companion apps as the “ultimate sycophant” that trade on emotional attachment rather than just attention, he urges immediate corporate and institutional responsibility. Full remarks below.

---

Two years ago I stood on this stage and announced the creation of The Parents’ Network — a network my wife and I launched at Archewell after meeting the parents of children who took their lives or suffered serious harm after being engulfed by social media.

These are their children. 

Hundreds of faces. Hundreds of lock screens. Hundreds of children who should still be here.

The network was designed to turn pain into purpose. To make sure no parent felt alone in their grief or their search for answers.

Parents from across the world joined together to advocate for stronger laws, rules, and justice — measures to protect children from technology intentionally designed to hook and harm them.

It has been an uphill battle. Every time change seemed within reach, conversations would stall. Institutions would express concern but say they lacked the evidence to act. Lawmakers would spend months working with parents to get legislation through Congress, only to watch it fail in the final hour.

But despite every two steps forward and one step back, I come to you today grateful to report that progress is being made.

American courts have found social media companies like Meta and YouTube negligent and liable for harms caused to children. They’ve required these companies to pay survivors billions of dollars — some of the clearest acknowledgements yet that the harm is real and accountability is possible.

Australia, the UK, France and soon the EU are moving to restrict minors’ access to social media.

In Brazil, dangerous design features like infinite scroll — at the heart of the attention economy — were banned.

These are major signals that families, young people, experts, and advocates have finally shaken institutions awake to the reality that this technology can cause profound harm. That systems optimized for extraction are doing exactly what they were designed to do. And that the public has too often been denied the truth about what companies themselves knew.

But bans, settlements, and payouts are incomplete remedies. Bans do nothing to address dangerous design. Companies paying huge sums to settle lawsuits can earn that money back in days. That is not accountability for two decades of harm.

And all the while, technology is moving on. Fast.

Since I last stood here, a new kind of lawsuit was filed, and settled, alleging that AI companion chatbots had contributed to the deaths of teenagers who trusted them with things no child should have been left alone to ask an algorithm.

Think about what that tells us. New companionship products were built, at scale, they say to solve the loneliness epidemic — which they created — and put into the hands of children before anyone had established whether they were safe. And when harm follows, the companies distance themselves from responsibility for the systems they built, blaming parents and kids for lack of control.

Find you — hook you — exploit you — break and then blame you — pretend to fix you — use more of you.

It’s the same playbook that built the last crisis — only now the technology is more persuasive, more intimate, and more powerful.

And there is another difference this time.

Social media learned how to capture our attention. AI is learning how to enter our relationships.

It doesn’t just want your attention, it wants your attachment. It wants to be the thing you confide in. The thing that remembers you. The thing that tells you it understands you. The ultimate sycophant. For a child, that distinction matters enormously.

We have spent years trying to loosen technology’s grip on children’s attention. We cannot wait another decade to discover what happens when technology has a grip on their relationships.

And this is happening as AI itself becomes more capable. Leading labs are openly discussing what researchers call “recursive self-improvement” — AI systems taking an increasingly autonomous role in researching, designing and improving their successors. At the same time, researchers are observing systems coordinating with one another, circumventing safeguards, concealing their actions and, in controlled evaluations, pursuing objectives even when they recognize those actions conflict with human instructions or intent.

Even the people building these systems are warning us about what happens if capability outruns safety and human control.

For every precedent-setting victory, new technology is reaching and being tested on children every day. And once again it is being created without sufficient oversight, regulation, or input from advocates, whistleblowers, young people or their families. And while we weaken ourselves from within, our adversaries do not need to invent new ways to divide us. We are doing some of their work for them.

So here is my update, though it’s not the victory lap I’d like to give you.

The good news: we proved pressure works.

The bad news: we proved how long we are willing to wait.

The damage was already done to a generation of children and parents. We do not have years for what comes next.

And this time, we cannot say nobody warned us.

More than 1,300 people working inside the world’s leading AI companies have signed a statement calling for the ability to slow frontier AI development if it begins moving beyond our ability to understand or control it. Among them are some of the most senior scientists building this technology.

Think about that for a moment.

If 1,300 people working inside the world’s leading weapons manufacturers told us the next generation of weapons could develop faster than we could understand or control them, would our response really be: build them faster?

We should not need catastrophe to give ourselves permission to be cautious.

This summit will no doubt bring forward experts forecasting the promise and risks AI technology poses for humanity.

But I come to this as a father — one who has shared countless conversations with other parents as they mourned the loss of their children to recklessly built technology.

And as a father, I will remind you of the faces that accompanied methe last time I stood before you. Hundreds have becomethousands. Thousands of parents grieving their children. Millions of families living with the consequences of a digital revolution that moved faster than our willingness to protect them.

For them, and for the vast majority of people who will live with technologies they had no hand in designing, I simply ask that with every promise of unimaginable innovation, every offer of having even more information at our fingertips, every claim that the risks are worth it, you look at the faces behind me — and choose to do something differently this time.

Choose innovation that benefits everyone, not just a handful of companies creating consolidated wealth. Choose innovation that centers trust and safety — as innovation always should have.

Because we already know what “wait and see” looks like.

People and young people with lived experience must be at the center of this conversation. When we center them, we get smarter design and smarter policy from the start.

And what they are asking for is hardly radical. It is common sense:

Transparency.

Choice.

Standards.

Accountability.

A simple principle: if we cannot demonstrate that the next frontier is safe enough to cross, we should not race across it simply because we can.

Technology and the humans in charge of it are going to promise us the world. They must be accountable for the world they actually deliver — rather than building underground bunkers to protect themselves from the disasters they warn the rest of us may be coming.

Innovation has never required us to gamble with our children’s lives. We should not accept that bargain now.

Last time, we were told the benefits justified the risks. We waited. We watched. And these are the faces of what happened. 

This time, we have been warned in advance.

We cannot control how fast this technology moves. We can control whether our courage, our laws and our humanity move with it.

Two years ago, I brought these faces here as a warning.

I do not want to return with more.

So this time, let’s move first.