The anniversary of September 11 invites remembrance before argument or advice. But from the distance of a quarter-century, and with thousands of suitably reverential pieces appearing around the world today, it is appropriate to ask what we have learned — or failed to learn — from this generation-defining psychological scar.
The mind still recoils from the horrors of that morning: 19 men boarded civilian airplanes intending to die, and to take as many innocent people with them as possible. It’s impossible for most of us to imagine dying gladly for the opportunity to murder scores of strangers. But we did not misconstrue the message of this death cult. The fear that gripped the nation was not irrational.
Nor were the calls for a federal response inherently wrong. A free society has both the right and the obligation to defend itself against people prepared to exploit its openness, turn its technology into weapons, and slaughter its citizens. A government unable or unwilling to protect its people from organized violence has failed at one of its few legitimate responsibilities.
When Defense Becomes Surrender
But the legitimacy of defending a free society can easily become a crusade to destroy the very freedoms it hopes to safeguard. As the crater in Manhattan was still smoldering, fear made demands for government restraint appear frivolous. Concerns about privacy seemed embarrassingly abstract beside smoking ruins and walls covered with photographs of the missing. Constitutional objections looked like the pedantic petulance of people who did not understand the danger civilization faced.
But this is the danger civilization faced — and faces still. When we are afraid — and human fear is not difficult to stoke — we demand that the powerful do something.
Congress passed the PATRIOT Act scarcely six weeks after the attacks: Uniting and Strengthening America by Providing Appropriate Tools Required to Intercept and Obstruct Terrorism (USA PATRIOT). Shortly after, new agencies were created and empowered to monitor Americans like never before; a workaround for warrant requirements was included. Under the guise of “stopping the next attack,” Americans were prompted to accept previously unthinkable expansion of surveillance authority (including the strategically but inaccurately titled ‘Department of Homeland Security’) and coercive powers: watchlists and no-fly-lists, bulk data collection, enhanced interrogation, secretive searches and closed judicial proceedings.
Every new intrusion came wrapped in an implicit promise: this isn’t for you, law-abiding citizen. This power will be aimed at them, the terrorists, the ones you’re afraid of. They don’t deserve privacy or dignity or human rights, the claim went, those fanatics who are plotting “the next 9/11” in mosques or libraries or encrypted chats. If you have nothing to hide, the familiar chant goes, you have nothing to fear.
To find those murderous needles, though, government authorized itself to collect the haystack. Millions of innocent people — most of them unaware — had their personal information, associations, financial transactions, private conversations, online activities, and other potentially useful information scraped and stored, subjected to a level of scrutiny that had rarely been mobilized even for a suspected serial killer before the PATRIOT Act passed. The distinction between being investigated and simply existing quietly collapsed. The notion of watching individual suspects became antiquated: the government could simply watch everyone, and store the results until they were needed.
The War on Terror’s legacy is still pushing forward, but it clearly contradicts the principle on which it was built: protecting a free society. If President George W. Bush was correct that terrorists “hate our freedoms,” then the 9/11 attack was largely successful: it induced us to indefinitely suspend constitutionally protected freedoms that Americans had long cherished.
Worse, the PATRIOT Act further entrenched a dangerous political habit: when fear and ambiguity arise, the answer is to make every citizen more visible to a less-accountable state.
The Next Emergency
Americans look likely to repeat the pattern as legitimate fears of AI’s destructive capacity grow. Automated cyberattacks, deepfakes, and never-before-seen digital powers will be available to anyone who wants to do harm.
Many who cite AI’s risks are sincere and well-informed. Others are manipulating our fears. But that distinction doesn’t matter, in the long run. Once an extraordinary threat is accepted, sincere fear is likely to authorize the same machinery as cynical pursuit of unchecked power.
Imagine that tomorrow you wake to find your bank account at zero. The news reports that major banks have been hacked by some AI agent or AI-assisted agents. Your savings can be restored to you — but only in this US-backed stablecoin, digital dollar, or central bank digital currency. And sure, the new system comes with more restrictions and precious little privacy, but would you care? The new security measures are only aimed at terrorists or AI agents anyway, so why should you object? If it saves even one life… The rhetoric is already familiar.
The perverted beauty of the AI threat is that the powerful can create exactly the pressures they seek to exploit.
There will always be a danger on the horizon. There will always be threats in the periphery that someone calculating can turn into a reason-defeating fear.
If AI becomes a civilization-defining threat, the way terrorism did, anyone who claims to be concerned about civil liberties will be caricatured as a denialist. Anyone who demands proof before government tightens the screws will be said to be gambling with human lives.
Uncertainty Becomes Permission
The arguments that “the consequences of inaction may be irreversible” and “by the time we have evidence, it may be too late” are extraordinarily powerful, not because they are true, but because they cannot be refuted. The hypothetical harm is used to justify the actual harm. Uncertainty itself becomes evidence that government power must grow, while liberty and privacy must shrink.
Each proposal would arrive alone, described as narrow, technical, and necessary. Each would target some unsympathetic actor: terrorists, foreign agents, cybercriminals, child predators, weapons developers, extremists. Each would promise to inconvenience law-abiding people as little as possible.
Unimaginably horrifying edge cases become the justification for the most minute micromanagement. Emergency powers transmute themselves into commonplace frustrations: can I take more than a few ounces of pumped breast milk through security? Can I visit family in my home country, even though my name is very similar to some Interpol suspect’s?
If the threat of AI weaponization becomes broadly accepted, the desire to escape government control could easily become evidence of criminal wrongdoing. Anonymity becomes inconvenient, then suspicious. Why don’t you want programmable money that can exclude any party the government disapproves of? Why do you object to PATRIOT Act powers upending your library records? What are you up to?
What Fear Can Destroy
Twenty-five years ago, Islamic terrorists sought to destroy far more than American buildings and American bodies. They hated and wanted to obliterate a society built on mutual tolerance, pluralism, liberalism, and the freedom of ordinary people to direct their own lives.
That glorious ethical tradition couldn’t be destroyed by weaponized airplanes. Nor will it be destroyed by AI agents attacking our power grids and personal finances.
What destroys free societies, ultimately, is fear. Fear of terrorist attacks, fear of what AI might do to us. Our own imaginations, overrun with input from limbic systems built for occasional attacks on the African savannah, will invent the tradeoffs that would have been self-evidently oppressive until recently.
But the most dangerous threats are exactly those that are realistic, but distant. A danger real enough to make every surrender of humanity and privacy seem quaint will result in genuinely formidable centralized control. Even if you happen to admire those who wield that power in the current administration, democracy and time virtually guarantee that the unchecked power we sometimes desire will be wielded by those we despise.
After September 11, too many people treated fear as permission, and uncertainty as an argument for control. The institutions proposed for that uncertain moment became infrastructure.
Twenty-five years later, with an even more amorphous enemy placed before us, we have another chance.
A free society must defend itself. It must also remain a society free enough to be worth defending.
