Ranking Member Raskin’s Remarks During Meeting of the Judicial Conference
Ranking Member Jamie Raskin
Prepared Remarks Before Chief Justice John Roberts and Assembled Jurists for the Opening of the Federal Judicial Conference October Session
September 17, 2026
At your Supreme Court confirmation hearing, a few decades ago, Mr. Chief Justice, you famously told our colleagues on the Senate Judiciary Committee: “Judges are like umpires. Umpires don’t make the rules, they apply them[,]” and you committed to remember on the bench that your job is only “to call balls and strikes, and not to pitch or bat.”
Well, Mr. Chief Justice, we live in interesting times, and it seems Artificial Intelligence may be coming for your job as an umpire at home plate. This year, Major League Baseball (MLB) introduced a new rule called the Automated Ball-Strike (ABS) Challenge system, which allows players to appeal an umpire’s “ball or strike” call. Powered by a multi-camera tracking system called Hawk-Eye technology, the new rule provides for precise real-time digital monitoring of each pitch in the game and allows for essentially infallible assessment of whether it was, in fact, a ball falling outside the strike zone or a strike traveling within it. Under the new rule, once the pitcher, the catcher or the batter signals a challenge to a call (they are the only ones who have the power to challenge and, even then, it must be within two seconds of the pitch), an animated graphic showing the exact trajectory of the ball relative to the automated strike zone appears immediately on the stadium videoboard and the live TV broadcast. The entire appellate process is visible, and it takes only about 15 seconds. Within a minute, the game resumes immediately.
This is literally a game-changer. Every first-year law student knows that Justice Robert Jackson wrote in a 1953 opinion: “We are not final because we are infallible, but we are infallible only because we are final.” So with all due respect to you, Mr. Chief Justice, to the late Justice Jackson, and to all of the fine judges who faithfully call “balls and strikes” for a living, the rulings of this new Hawk-Eye-powered ABS Challenge System are not infallible only because they’re final. They are also infallible because they are actually infallible (at least to a distance of within 1/6th of an inch, according to MLB). Because they are overwhelmingly infallible, they do become final.
Now, under the new rule, the human umpire still calls all balls and strikes, and the ABS Challenge System limits each team to just two challenges to the ump’s call over the course of a 9-inning game. If you use a challenge (what you guys would call an appeal) to question a ball or strike, and you win on it, you get the call on the field reversed and you get to keep the challenge going forward. If your appeal is incorrect and therefore fails, you lose not only the call but that challenge. Thus, petitioners, beware—use your appeals carefully and sparingly.
Ever since the new baseball rule came into being, I have been thinking back to your confirmation hearing and worrying about whether your own job might be similarly threatened one day by the legal equivalent of the Automatic Ball-Strike System. Could an AI-driven system which digests the entire case record, all relevant or cited case authority and all relevant legal, statutory or constitutional texts, and deliver in a matter of seconds the perfectly mechanical, objective and logical application of the relevant law to the relevant facts actually give us the ultimate, infallible appellate robot Justice to call judicial balls and strikes?
The question, I think, immediately points to the inadequacy and insufficiency of the “calling balls and strikes” metaphor as it applies to appellate adjudication. For example, for many Justices, judges, lawyers and citizens, the constitutional challenge by civil rights groups to President Trump’s effort to nullify birthright citizenship in America by executive order was a perfect strike thrown right down the middle. After all, the first sentence of the Fourteenth Amendment reads: “All persons born or naturalized in the United States, and subject to the jurisdiction thereof, are citizens of the United States. . .” A plain reading of the text of the Constitution (textualism), a searching review of the history and original understanding of the Fourteenth Amendment (originalism) and a study of all prior decisions on the issue (precedentialism) all seemed to settle the matter definitively.
But even in that decision, which appears to be an open-and-shut case of a pitch barreling down the center of the strike zone, there is more to the picture than meets the eye. The three main dissenters in the case, Justices Thomas, Alito and Gorsuch, gave a completely different gloss to the Citizenship Clause, contending that it does not apply to children born to foreign nationals and arguing that the history should be read narrowly to mean that the Fourteenth Amendment was designed to apply to the children of enslaved people, not the children of foreigners.
In other words, Justices with equal access to all the legal materials came up with completely different answers on about the most straightforward constitutional question you can imagine.
The issue is that appellate adjudication does not just involve factually calling balls and strikes (a lower-court process which is, to be sure, sometimes vexed and indeterminate in itself) but actually interpreting the rules of the game and defining their meaning for future cases. On the Supreme Court, you are not just calling balls and strikes but applying different modes of interpretation to define and elucidate what the rules actually are.
When I visited you last year, I discussed the Steel Seizure case (1952), where the Justices interpreted the Constitution and drew from it a rule: that when the President acts without an explicit grant of power from the Constitution or a specific Congressional authorization, his authority is at its lowest possible point. The Court arrived at that rule through textualism, originalism, and constitutional structuralism, but there were several dissenting opinions expressing disagreement from a few justices who wanted to elevate executive power over those constitutional methodologies. All of which is to say that there are always different constitutional values and political purposes informing the thorough human process of discerning and applying the meaning of the Constitution. While the umpire calling balls and strikes is only applying the rules set forth in Major League Baseball’s Official Baseball Rules, the Judiciary is applying and interpreting the broader and infinitely more complex language and contested meanings of the Constitution and federal law. We must concede that Constitutions and laws are born of politics and never really leave the field of moral and political contest.
Perhaps an AI robot could perform mechanical application of well-settled rules to simple facts—but even then, it would work only until some human Justices showed up and proposed a complete paradigm shift or interpretive realignment in understanding the Constitution.
So, Mr. Chief Justice, the good news is, I don’t think you have too much to fear from AI -robot Justices taking over your job. Our judiciary depends on human judges who not only call balls and strikes on the facts but develop rules where they have not yet been articulated and render judgments that consider not just the facts and the law, but also the equities of each case. Perhaps the smartest AI program could set forth the rival interpretive choices and consequences implicated in any act of adjudication but how could it choose among all the different options?
At the end of the day, the character of the judge (or the Justice) is essential, indispensable and ineradicable. The judicial character simply cannot be reduced to an AI program whose values have never been tested in life, much less confirmed by the United States Senate.
But let us close with an important implication of this analysis. Just as human character supplies the value-based decision-making ingredient that AI does not have, human beings have a flaw that AI, at least up until now, has not displayed: personal bias and prejudice in our own favor. This is why the whole history of jurisprudence—going back to Roman Law, Sir Edward Coke’s meditations on English Common Law and John Locke’s philosophical underpinnings for civil justice—influenced James Madison to enunciate in Federalist No. 10 the cardinal principle that anchors every part of our justice system: “[n]o man is allowed to be a judge in his own cause, because his interest would certainly bias his judgement, and, not improbably, corrupt his integrity.”
Thus, the Supreme Court, like all other courts in the land, needs an ethics system where allegations of a Justice’s disqualifying conflicts of interest in a case or general ethical lapses will be addressed and resolved by a process wholly independent of that Justice’s will. This is the way it works in every other court in the land, including 50 state Supreme Courts. While we may certainly hope that all our Justices will be wise in rendering decisions in other people’s cases, we cannot presume that any of them will be superhuman and objective in deciding their own.
If we don’t want future generations to argue for the complete hand-off of appellate adjudication to unbiased and impartial AI robots and programs, we had better get serious about applying this essential Madisonian principle to the operation of our own courts right now. That is a call that any fair umpire anywhere would have to make.