Directionally Correct

Fractionally Legal 63 Looks at Having the Right Idea and the Wrong Map

“Directionally correct” is what I tell a client when they have the right idea and no map to get there. Or, more often, when they are holding the map upside down. I have been seeing a lot of that lately, in our politics and in the technology du jour, which is AI. The instinct points the right way. The execution wanders into a ditch.

First, the Trump administration. Even where Trump has been directionally correct, the incompetence, the corruption, and the politicization get in the way. The map is upside down and also, somehow, still in the glove compartment (or your phone with Google Maps is powered off). I happen to think Tariffs are a good idea – directionally correct – but they were a disaster because of their randomness, the endless exemptions, and because Trump never bothered to build a case or get them through Congress. The immigration crackdown, which most Americans think is directionally correct, has been cruel and violent, with no real plan to bring anyone in except, for reasons I still cannot explain, a handful of white South Africans. Very few Americans think the war in Iran was directionally correct, but assuming it was, it has been a disaster in roughly every direction a war can be a disaster.

Which brings me to the new “Anti-Weaponization Fund.” Right idea. Upside-down map.

The impulse behind the fund is correct. People who are subjected to politically motivated government prosecutions should be compensated. I have written before about the hole in our law where that remedy ought to be. We have a statute, Section 1983, that lets you sue state officials who violate your constitutional rights. We have nothing equivalent for federal officials, because the Supreme Court keeps narrowing Bivens to the vanishing point (most recently in Goldey v. Fields last June). I laid all of this out in Fractionally Legal 53, in the context of the killing of Renee Goode by a federal agent who the administration insists has “absolute immunity.” So if you had told me the federal government wanted to build a real mechanism to make citizens whole when its own officers trample their rights, I would have been the first to cheer.

That is not what this is.

The $1.78 billion Anti-Weaponization Fund is blocked for the moment. Trump effectively settled a lawsuit with his own administration and then proposed to fund the payouts out of the federal Judgment Fund, which is unlawful precisely because the president sat on both sides of the table, as plaintiff and as head of the executive branch he was suing. You can’t make this stuff up.

And who is the fund for? Trump’s allies who claim they were targeted in the prior administration. Pressed on whether people convicted of assaulting police officers on January 6 could collect, Acting Attorney General Todd Blanche would not rule it out: anyone can apply if “they believe they were a victim of weaponization.” Sure.

If you actually wanted to compensate the victims of politicized prosecution, the first names on the list would be the people this Justice Department went after: Jim Comey and Letitia James, whose indictments were dismissed, and now E. Jean Carroll, and of course the immigrants detained in circumstances where a warrant was required and none was obtained. Instead the fund points the other way, at Biden-era grievances, which tells you it was never about the principle. It was about the payees.

Two structural problems doom it on top of the politics. First, it never went through Congress. Compare that to the one large government compensation fund I know intimately. It was actually a precedent that Trump’s people pointed to when setting up the new Anti-Weaponization Fund: the Black Farmers discrimination litigation. I know it well because I was counsel to an intervener in that long-running litigation. That settlement was funded because Congress appropriated the money through the Claims Resolution Act of 2010. It was court-supervised, it had a trustee, and it had real restrictions on who got paid and why.

Second, there is no attempt to help anyone outside Trump’s favored circle. A fund that only compensates one side’s political friends is not a remedy. It is a slush fund.

So here is my directionally-correct counter. Litigate better. If the goal is justice, give us a real cause of action against federal officers who violate constitutional rights, the federal analogue to Section 1983 that Congress has refused to pass and the Court refuses to invent. That would not be a politically motivated production. It would be a real solution. Right idea. This time, hold the map right side up.

Now to something else where we are directionally correct and still cannot read the map. We know AI can do legal work. I have written about this many times: I am all in on using AI to serve my clients faster and cheaper, and I do it every day. What almost nobody has figured out is the rest of it. What should we use it for, what should we keep it away from, and how does it fit with the way lawyers actually work?

I think there are three models at the moment. Two are directionally correct but hopelessly lost. One offers a real map.

Model 1: The consultants. This is Harvey, Legora, and the hundreds of thousands of “legal agents” the marketing decks promise. They are not practicing law. They are selling expensive software to the lawyers who do. Two problems. The first is cost: these are pricey, subscription-heavy products, and most of them are a solution hunting for a problem, Frankenstein’s monster looking for a soul. The second is the tell I keep coming back to: the lawyers with the most money to spend on these tools are the ones who need them least, because their clients will happily pay for human bodies. If your matter is a multi-billion-dollar deal or the trial of the century, money is no object and AI is barely relevant. You are better off, frankly, with high-priced humans, and that work is not going anywhere. Also, these tools run on LLMs that are wildly expensive to operate at scale, and the whole experiment is currently underwritten by investor money. When that subsidy ends, so does the math.

Model 2: The black box, also sold as the “neo” law firm. These are the platforms that try to route around the lawyer entirely. In my experience they are one of two things. Sometimes it is slick marketing by lawyers who are doing perfectly ordinary legal work and dressing it up as a robot for reasons I don’t really understand (it probably has something to do with raising money from outside investors, which is of questionable legality). Other times it is a genuine, longer effort to deprecate the lawyer out of the picture as the technology improves. Either way, the limitations are structural, not cosmetic. Only a licensed human can practice law. The AI cannot appear in court, cannot exercise judgment, and cannot be disciplined when it screws up. Which means until the lawyer really can be deprecated away, the black-box model is charging you for the AI and the lawyer. You are paying twice for a service sold as cheaper. This model may yet win, but it’s a long way off.

Model 3: The AI-augmented lawyer. This is where I live, and I think it is the right answer, at least until the economics and the responsibility question get sorted out. A licensed lawyer uses AI as a force multiplier. Faster drafting, better document review, quicker research. But the lawyer does the actual lawyering, which is judgment, and stays on the hook for it. I draft with AI constantly but no one is really paying me for the draft anyhow. I’m paid for solid legal work and advice that allows them to run their businesses better and mitigate risk.

I think fractional general counsel is the ideal expression of the augmented model, because the whole pitch of fractional GC work is faster and cheaper. The value of a GC was never the contract she drafts or the brief he files. It is the judgment about whether to draft the contract at all, and what it needs to say. That is the difference between a tactical lawyer and a strategic one. AI makes the tactics almost trivial which frees up the strategy, which is the part you are actually paying a GC for. AI handles the mechanics. The lawyer handles the judgment. The client gets a strategic advisor at a fraction of an in-house GC’s cost or a law firm, faster than traditional outside counsel. And I’ll still answers the phone when it matters.

That is the whole thesis behind YFGC.AI. I am not trying to be an AI law firm. I am trying to be a better lawyer. AI is how I get there. Right idea. And I’m holding the map right side up.

Keep thinking, keep building.

Jesse

Hi, and welcome to my newsletter. I’m Jesse Strauss, Your Fractional General Counsel. I’m a lawyer with a private practice based in New York City, helping clients in the United States and globally with their U.S. legal needs. My expertise spans raising funding rounds, employment issues, negotiating master service agreements, intellectual property, compliance, legal process management, and dispute resolution. My focus is on founding and nurturing great companies from seed to exit.

Learn more at YFGC.AI and book a complimentary 30-minute consultation. You can also subscribe to my Substack and follow me on LinkedIn.