In late September 2026, Donald Trump and a gathering of technology executives signed what they called an artificial intelligence constitution — a voluntary framework of principles meant to govern how AI is developed and deployed, chosen deliberately over mandatory federal regulation. The agreement reflects a recurring tension in democratic governance: whether emerging powers are best tamed by those who wield them or by those who must live with their consequences. On the same day, former special counsel Jack Smith testified before the Senate, a reminder that accountability for the past and stew
Trump, Tech Leaders Sign AI 'Constitution' as Smith Testifies Before Senate
A promise that depended entirely on goodwill and business judgment
Mark
So what exactly did they sign? Is this a binding agreement?
Mimi
It's more of a commitment to principles. Companies agreed to certain practices around testing and disclosure, but there's no enforcement mechanism—no penalties if someone doesn't comply.
Luke
That's worth flagging. We don't know yet whether these are meaningful commitments or just cover. The source material doesn't tell us which companies signed on, what the actual standards are, or whether there's any way to verify compliance.
Mark
Why would Trump's team prefer voluntary controls over regulation?
Mimi
It fits with a deregulatory philosophy. Industry-led standards let companies shape the rules before government mandates them. It's also faster than legislative process.
Luke
But we should be honest: we don't have detail on what the administration's actual position is, or whether this was their idea or the tech industry's pitch.
Mark
And Jack Smith's testimony—what was that about?
Mimi
He appeared before the Senate on the same day. It suggests ongoing federal investigations and scrutiny, even as the political landscape shifts.
Luke
The source material is thin here. We know he testified, but we don't know what he said, what he was asked, or what the implications are. It's a fact, but it's incomplete.
Mark
So we're looking at two different approaches to accountability—one forward-looking, one backward-looking?
Mimi
Exactly. One is about preventing future harms through cooperation. The other is about examining past conduct through formal process.
Luke
And the real question is whether the voluntary framework will actually work, or whether it'll collapse under pressure and force Congress to act.
O Pulso
- With AI capabilities accelerating and public concern mounting, Washington faced urgent pressure to establish rules before the technology outpaced any meaningful oversight.
- Tech leaders moved first — securing a voluntary framework that lets them write their own standards rather than absorb mandates from legislators who may understand the risks but not the architecture.
- The resulting 'constitution' is built entirely on goodwill: no enforcement mechanisms, no penalties, no independent auditors — just a collective promise from the very companies with the most to gain from lenient terms.
- Across Capitol Hill that same day, Jack Smith's Senate testimony kept alive the question of federal accountability, signaling that investigative scrutiny does not simply dissolve when administrations change.
- The voluntary approach now faces its proving ground — if tech companies treat the document as a branding exercise rather than a binding commitment, congressional appetite for mandatory regulation stands ready to fill the void.
In late September 2026, Donald Trump and a gathering of technology executives signed what they called an artificial intelligence constitution — a voluntary framework of principles meant to govern how AI is developed and deployed, chosen deliberately over mandatory federal regulation. The agreement reflects a recurring tension in democratic governance: whether emerging powers are best tamed by those who wield them or by those who must live with their consequences. On the same day, former special counsel Jack Smith testified before the Senate, a reminder that accountability for the past and stewardship of the future often share the same calendar. Together, these two events trace the outline of a government searching for its footing in an age that is moving faster than its institutions.
On a Tuesday in late September 2026, Donald Trump and a cohort of technology executives formalized what they called an artificial intelligence constitution — a set of voluntary standards governing how companies would build and deploy AI systems. The choice of voluntary over mandatory was itself the message: the industry and the incoming administration preferred self-governance to federal prescription, consistent with Trump's broader deregulatory posture.
The document outlined principles — commitments around testing, disclosure, and risk assessment — but carried no enforcement mechanisms, no penalties for noncompliance, and no independent oversight with real authority. It was, at its core, a promise, and one whose value would depend entirely on the integrity of those who made it.
The timing was deliberate. As concerns about AI misuse, bias, and unchecked deployment intensified, tech leaders moved to shape the regulatory conversation before stricter mandates could take hold. The voluntary framework offered them that window — but also set a clock. If companies failed to demonstrate genuine self-regulation in the months ahead, Congress, which had already shown appetite for harder rules, would have every reason to act.
Meanwhile, across Capitol Hill, Jack Smith testified before the Senate about matters under federal investigation, a counterpoint to the forward-looking optimism of the AI signing. His appearance was a reminder that accountability for past conduct and governance of future technology were both pressing concerns — and that Washington, in the fall of 2026, was being asked to manage both at once, in an era of rapid change and deep institutional mistrust.
On a Tuesday in late September, Donald Trump and a collection of technology executives gathered to formalize what they called an artificial intelligence constitution—a document outlining voluntary standards for how companies would develop and deploy AI systems. The agreement represented a deliberate choice: rather than submit to mandatory federal regulation, the industry and the incoming administration opted for a framework built on self-governance, with companies pledging to adopt certain safeguards and transparency measures of their own accord.
The timing was significant. As AI capabilities accelerated and concerns about misuse, bias, and unchecked deployment mounted, pressure had been building for Washington to establish clear rules. The voluntary approach offered tech leaders a way to shape the conversation before stricter mandates could take hold. It also signaled the Trump administration's preference for industry-led solutions over prescriptive government oversight—a posture consistent with the incoming president's broader deregulatory agenda.
The constitution itself outlined principles rather than hard requirements. Companies would commit to certain practices around testing, disclosure, and risk assessment. There were no enforcement mechanisms built in, no penalties for noncompliance, no independent auditors with teeth. The document was, in essence, a promise—one that depended entirely on the goodwill and business judgment of the signatories.
On the same day, across Capitol Hill, Jack Smith appeared before the Senate to testify about matters under federal investigation and legal scrutiny. Smith, whose office had been central to several high-profile cases, faced questions about the scope and conduct of ongoing inquiries. His testimony suggested that despite shifts in political winds and changes in administration, certain investigative threads remained active and subject to congressional attention.
The two events—one looking forward to a new framework for emerging technology, the other examining the conduct of federal law enforcement—illustrated the competing currents running through Washington in the fall of 2026. One was about preventing future harms through voluntary cooperation. The other was about accountability for past actions through formal process. Together, they reflected a government grappling with how to manage both innovation and oversight in an era of rapid technological change and deep institutional mistrust.
The voluntary AI constitution would face its first real test in the months ahead. Tech companies would need to demonstrate that self-regulation could actually work—that they would enforce their own standards, that transparency would be genuine, that the framework would adapt as new risks emerged. If it failed, if companies treated the constitution as a public relations gesture while continuing business as usual, pressure for mandatory regulation would almost certainly intensify. Congress had already shown appetite for stricter AI oversight in previous sessions, and the voluntary approach would need to deliver tangible results to forestall that push.
Quer a matéria completa?
Leia o original em CBS News ↗