AI Doesn’t Just Assist Decisions Anymore. It Shapes Them Before You Ever Get to the Table.
In most of my work, people assume the negotiation starts when the conversation starts.
It doesn’t.
By the time you’re in the room, whether it’s a deal, a dispute, or a strategic partnership;most of the important shaping has already happened somewhere else.
Increasingly, that “somewhere else” is an AI system.
Over the years, I’ve sat in different seats: tech founder, CIO, CTO, general counsel, and advisor to startups and institutions. What’s changed recently is not just the speed of technology. It’s the fact that decisions are being pre-structured before any human formally engages them.
Risk scoring has already happened.
Visibility has already been filtered.
Options have already been ranked.
And no one in the room can always explain exactly how.
We’ve already seen versions of this play out in real life.
In State v. Loomis, a risk scoring algorithm influenced sentencing decisions in ways even judges acknowledged they could not fully interrogate. In hiring systems at major tech companies, AI tools have quietly filtered candidates before a human ever saw the application. In credit decisions like the Apple Card controversy, people questioned whether algorithmic outputs were producing outcomes that even the institutions themselves struggled to fully explain or justify.
Different industries. Same pattern.
The system shapes the starting point before the human conversation begins.
That creates a very different kind of negotiation environment.
Because you are no longer negotiating just with the counterparty.
You are negotiating inside a system that has already influenced what both sides believe is possible.
From a legal perspective, this is where things get interesting. The traditional framework assumes you can trace decisions back to intent, instruction, or oversight. When AI systems are embedded in workflows, the decision becomes the output of multiple layers of optimization rather than a single point of judgment.
Which raises a practical question I find myself thinking about more often:
When leverage is shaped upstream by systems you don’t fully control, where does accountability actually sit?
Not in theory. Not in abstraction. In practice.
Because in real deals, what matters is not whether the system is “right” or “wrong.” It is that it quietly shifts the starting point of negotiation without anyone explicitly agreeing to that shift.
That shows up everywhere:
In how opportunities are surfaced—or never surfaced at all.
In how counterparties are pre-qualified before human conversation.
In how internal recommendations are framed before leadership even weighs in.
This is where my worlds now overlap more than they used to.
Dealmaking, technology, and law are no longer separate lanes. They’re interacting inside the same decision environment.
And the skill that matters most in that environment is not just understanding what the system says.
It’s understanding what the system has already decided for you before you ever respond.
That’s the part most people don’t see.
And it’s where leverage is quietly moving.