Book Review: Clegg, N., (2025), How to Save the Internet, Bodley Head/Penguin

Until January 2025, the author, Nick Clegg, was a senior officer at Meta for seven years. Before that, he was Deputy Prime Minister in the UK and Chairman of the UK’s Liberal Party. Mr. Clegg focuses on generative AI. As generative AI is seen to reshape everything from communication to geopolitics, the book raises a major dilemma: Will open, democratic values shape the future, or will closed authoritarian models win? The global internet, traditionally open, may be at risk of becoming more closed, in line with the shift towards a seemingly more autocratic world.

The author argues strongly that in our democracy-led world, we should resist the temptation to turn the internet into a closed entity.

This book speaks about power, and the dilemma power raises for many of us. While we generally welcome the benefits of technology, we tend to resist the evidence or idea that power in the hands of a few can actually be risky, even dangerous. Many tech leaders have lined up with Trump’s America First philosophy, betting on the hope that perceived competition with countries such as China will leave these firms largely unregulated and thus still competitive. Yet this may lead to more closeness, away from its open global, very nature.

Mr. Clegg argues that it will be in the large tech companies’ own best interest to become more open. If not, these firms may be facing regulatory constraints in the future, especially under new political regimes. The author argues that Silicon Valley may be contributing to this closeness. The valley hosts a highly educated and talented population, with plenty of creativity and no limits to ambition. An engineering mind seems to prevail, namely, to find solutions to problems and then to move on to something new, something more innovative. There is little focus on the above-mentioned dilemma – open versus closed – however. The prevalent engineering optic does not leave much room for this.

Mr. Clegg offers interesting insights into Meta’s leader and major owner, Mark Zuckerberg, as well as the firm’s former CEO, Sheryl Sandberg. They are both strong drivers with an appetite to keep learning, and with generally humble minds.

Now, back to one of the central themes of the book. Why did Mr. Clegg decide to join Meta? He offers three reasons:

-       Optimism regarding technology

-       Further clarification of how tech firms and governments might interact

-       Goal to find new ways to build bridges between technology and politics

One major issue is the rise of ake news, hate speech, online targeting of political opponents and so on. But is the internet to be blamed for this? The author does not think so. Some forms of regulation might be found, however, but he suggests that social responsibility should be put alongside profit-making motives, and that innovations might be channeled in the direction of solving the problems identified.

A main determinant for how technology firms may be increasing in size is the quality of the platforms that these firms have. The “big time” US-based firms (Alphabet, Apple, Amazon, Microsoft and Meta) are now very much in the AI race. It seems powers of their respective platforms is a critical element.

Countries such as China and Russia are also in the AI race. Their models appear more closed than what we find in the US. The approaches taken in these two countries also give their governments the power to clamp down on political opponents. Speech plays a particularly important role when it comes to this. The closed internet has been purposely developed to restrict speech.

How do firms such as Meta regulate, at times, difficult content issues? Meta has established an independent Oversight Board, but there will always be borderline cases where the right course of action stays a matter of debate and judgment. It should be noted that this Oversight Board ruled to suspend President Trump from using the internet. The issue of removing political content can often be difficult.

The introduction of Chat GPT by OpenAI, where Microsoft is a major shareholder, was a shock to the tech industry. Do such generative AI platforms diminish AI-based research with focus on other areas, such as medicine (diagnosis, therapy), aviation carbon emissions, weather pattern prediction, solar and planetary images, or, even deciphering ancient scrolls. But Chat GPT and other generative AI platforms seem to have led to a growing fear that authoritarian regimes could use these to initiate cyber-attacks, and/or even to attack people. Generative AI has the potential to be both sword and shield! These systems “learn” over time, which may lead to a loss of human control.

Taking this one step further, we might soon anticipate cases where AI might influence us n ways that we likewise might not have followed when it comes to reactions to other humans. In the end, we should keep in mind that it is our actions, choices and preferences that dictate what the systems do.

To now focus on AI’s upsides. There seems to be no doubt that AI can contribute towards solving the productivity puzzle. Technology will replace certain skills, typically those that are basic and repetitive. Other skills become more highly prized. Gained productivity!

Are we seeing an end to the internet’s golden age? To better understand this question, one might assess how data is now increasingly being stored. While in the past, there was a tendency to store data anywhere where it would be safe, i.e., in line with an open internet, there is a growing trend towards storing data in each country, so-called “data localization”.

Paradoxically, this trend towards data localization, often seen as part of the race to become or remain number one in AI, could then often lead to more national protectionism. Clegg claims that it is the race with China to stay on top of AI that has contributed to increased protectionism. Here, it may be interesting to consider if the US will create a vacuum. Who will fill it? China is perhaps the most likely candidate, but can the US rally some of the world’s other techno-democracies, such as the EU or India, to fill the gap?

The author claims that major tech companies can take action, often significant, to counter this:

-       Open up the books

-       Develop decision-making power, not by focusing just on products and emerging technologies, but on new forms of governance too

-       Bring users into the process

-       Give users meaningful control

There are uncomfortable truths in this discussion, both for tech companies as well as for politicians. Tech firms do not typically live up to the four issues delineated in the previous paragraph. Politicians are falling behind when it becomes necessary to legislate to put and resolve any of the above.

Clegg sees openness as central to “saving” the internet, both an open internet, as well as sharing AI. And there could be many upsides, not only for the US, but for the EU and India as well!

It is perhaps important to ask whether a non-technical individual such as Nick Clegg is in the position to write about the internet and generative AI. In this reviewer’s opinion, the answer is a clear yes. This lack of technical lingo, often found in many other books on this topic, is refreshing! The book is easy to read, not least because of its uncomplicated language. There has clearly been major research coducted for this book, evidenced by the many excellent expert quotes that appear throughout the book. An absolute must-read!