Whither Tech Policy?

Some lunchtime remarks.

Photo by Theodore Poncet on Unsplash

Last month, I gave some remarks at a technology-policy summit hosted by the inimitable James Madison Institute. The prepared text, edited during delivery as the clock expired, was as follows:

---

First off, I want to thank the James Madison Institute for having me — and, more importantly, all of you — here to mull over the development of the ideas and policies that shape and govern American technology.

We find ourselves in the middle of, shall we say, an energetic period of thinking, writing, and legislating on tech. And well we should be: as in the 19th century the Industrial Revolution rearranged the very foundations of the economies and societies that mechanized, the Digital Revolution today is changing some of the fundamentals of innovation, communication, and social relationships.

In Gutenberg’s day, the printing press disseminated information further and faster than any other previous invention ever had, which caused nervous onlookers to denounce it as a purveyor of disinformation — as an uncontrollable force, in equal parts powerful and dangerous, that would coarsen intellectual life. They are echoed today by their successors, who denounce artificial intelligence and social media on many of the same grounds.

Neither the Industrial nor the Digital Revolution resulted only from technological advances. Their essential source was freedom — the freedom to discover, the freedom to experiment, the freedom to associate, the freedom to do things differently from everyone else.

The American tech sector was conceived in liberty, and it has become the greatest tech sector in the world, our national champion industry. However, great change often creates great anxieties, and anxious lawmakers both in statehouses and in Congress seem ever more determined to seize some sort of control of digital innovation and digital platforms.

Recently, however, regulators’ benign neglect and the intrepid spirit of permissionless innovation from which the free internet sprung seems to be going out of vogue. Many lawmakers now seek to manage digital technologies and the online lives of those who use them. To illustrate, consider this fact: in 2025 alone, on the topic of artificial intelligence alone, more than 1,000 bills appeared in state legislatures.

Now, many of these proposals are useful and necessary; the law must, of course, be adapted to changing circumstances. But many others proceed from misguided notions about technology and faulty conceptions of how digital life ought to be regulated.

Perhaps the fiercest debates in recent years have raged on the front of free speech and free expression. Lawmakers have put forward bills to control what is said online, what isn’t said online, who may speak online, how online platforms sort and promote and moderate speech, how artificial intelligence generate speech, and so on.

A recurrent theme in this discordant symphony is mandated age verification, a proposal devastating to users’ privacy and cybersecurity. Perhaps worse, today’s age-verification proposals amount to a show-your-papers regime, conditioning users’ exercise of their First Amendment rights in the digital world on their willingness to produce government-issued papers or biometrics.

The popularization of artificial intelligence has also proven especially discomforting to some. People, being people, like to say ridiculous and offensive things, and now they have begun to use AI to do so.

Seeing this, anxious regulators blame the AI models themselves and seek regulatory controls to prevent AI from producing objectionable outputs. This attempt rests on a bad premise, however: As long as people are determined to say objectionable things — and by that I mean things that officials believe subjectively to be objectionable — they will say them, AI or no. The problem, insofar as a problem does in fact exist, isn’t the technology, but human nature.

It's a simple fact that any attempt to root out the misuse of AI completely must, in order to succeed, be so draconian as to destroy the utility of the AI and all of its benefits for free speech altogether.

One of the arguments advanced for these kinds of speech regulation is that the troubles created by new technologies and digital platforms are so severe that something, anything, must be done to combat them. Many simply feel that they must rush to support ill-crafted bills which with more time and reflection they would likely oppose.

Others, feeling political momentum turning towards regulation, have come to believe that since something is likely to be enacted we’d better find the least flawed proposal and support it.

Both have made their peace with legislation that — for expediency’s sake — infringes in some degree upon the First Amendment. And since we are attending a James Madison Institute conference, I think the words of the man himself — the Father of the Constitution — might be of use.

Madison wrote in 1792 that “Conscience is the most sacred of all property.” The person’s right to a free conscience cannot be severed from its cognate right to free speech. It isn’t insignificant that the Greek word logos means both “reason” and“speech.” The two simply can’t be severed. As Madison argued, “the public faith is pledged” to defend the right to conscience, “by the very nature and original conditions of the social pact.” That is, free conscience and free speech are things upon which free governments, by their very nature, may not intrude.

In opposing the censorial Sedition Act of 1798, Madison called the right to free speech “the only effectual guardian of every other right.” He noted further that Virginia’s convention had, in ratifying the Constitution, declared, “that among other essential rights, the liberty of conscience and of the press cannot be cancelled, abridged, restrained or modified by any authority of the United States.”

I’ll return for a moment to the modern argument that something simply must be done about the changes wrought by new technologies, an argument often used to whip support for small infringements upon the First Amendment for the purpose of fending off larger ones. In effect, the argument is that we should do something bad to avoid something worse.

Madison, I think, would have none of this. Securing free speech is one of the fundamental ends of the American experiment, and free speech is one of the indispensable ingredients of healthy and productive politics.

Madison saw that the principles of governance we choose to establish have far-reaching consequences. He observed this of the colonists’ revolt against Great Britain: “The free men of America did not wait till the usurped power had strengthened itself by exercise, and entangled the question in precedents. They saw all the consequences in the principle, and they avoided the consequences by denying the principle.”

And as I see it, there are two mutually exclusive principles now struggling for supremacy within tech policy. On the one hand, there is the principle that says that the traditional free-speech protections that obtain offline must not be curtailed with respect to online speech; on the other, there is the principle that says that digital technologies are different from offline technologies and must not receive full First Amendment protections.

The consequences of the principle that First Amendment protections need not apply fully online will not end with whatever legislative compromises on free speech are made today. Bending the First Amendment online would destroy its very structural integrity. More restrictions of speech will follow in future years. If we permit microfractures to be introduced into our now rock-solid free-speech law and jurisprudence, those microfractures will eventually widen into far larger cracks.

I think the distinction between “the digital world” and “the real world” is often overstated. They are each a part of the same world, inhabited by all of us. Constitutional safeguards and knowledge problems apply to both. Allowing even the smallest incursions upon online free speech simply because that speech occurs on a social media platform, and not in a newspaper or over the airwaves, would invite the regulators to exert ever tighter control over the conscience of a nation.

But what of the despicable online speech? Speech that is illegal offline should not be permitted online — and it is not. True threats, calls to imminent lawless action, and defamation are no more legal on Twitter than they are anywhere else.

As for speech often colloquially called “lawful but awful,” that speech is constitutionally protected irrespective of where it occurs. A great moment in American history was the Supreme Court’s decision in Brandenburg v. Ohio, in which the ACLU succeeded in defending the speech rights of Nazis. We as a nation pride ourselves on Brandenburg not because we approve of that most horrific of ideologies, but because the exiling the government entirely from the business of policing speech is one of the great achievements in the history of freedom. That achievement must not be allowed to slip away.

Another case brought by the ACLU, ACLU v. Reno, made clear that all the ordinary protections of the First Amendment do, in fact, apply to online speech. And in yet another ACLU case, 2004’s Ashcroft v. ACLU, the Supreme Court made clear that online age verification imposes constitutionally impermissible burdens on online speech. Since 2004 courts have routinely ruled unconstitutional laws attempting to mandate age verification, control platforms’ content-moderation policies, and suppress disfavored speech. These are faithful interpretations of the First Amendment and its jurisprudence.

Although, as I said, new technologies usually excite new anxieties, we must continue not just to remember but safeguard the cornerstone principles of American liberty, both online and off-.

The rise of digital technologies present both a blessing and curse. The blessing is that on the internet, information can travel faster and freer than ever before; the curse is that information traveling over digital pathways can readily be stifled and manipulated by small-minded government officials.

Technology liberates, but if we allow it to be micromanaged by the state, it threatens to enable authoritarianism to a degree never before known. Conversations that once occurred in private between friends at a bar or across a living room now occur over digital networks, where censorial regulators can snoop and intervene — but only if we allow them to do it.

Today, in tech policy, we find ourselves tasked with the eternal project of defenders of liberty: to stand athwart history, yelling Stop — even “at a time,” in the words of Bill Buckley, “when no one is inclined to do so, or to have much patience with those who so urge.” Now, as never before, we must commit ourselves to Barry Goldwater’s maxim: “extremism in the defense of liberty is no vice.”

For all that I’ve emphasized the inviolability of free speech, I think it’s fair to say that if I were to delay you all any longer in getting to lunch, you’d be entirely justified in having me locked up.

So, to close, I’d like once more to thank the James Madison Institute for hosting us and thank you all for the work you do every day to safeguard the free and open internet for generations of Americans to come. Thank you.