United States v. Heppner and the privilege gap

What people actually do

Millions of people discuss legal problems with AI every day. Not because they don't take their situations seriously — because they do, and AI is the least intimidating way to start.

Someone gets a termination letter at 5pm on a Friday. Scared, angry, confused. Calling a lawyer feels like a big step — you don't know the cost, can't explain your situation, and don't want to sound stupid. So you open a chatbot and type: "My employer fired me today. Can they do that?"

That conversation feels private. It's just you and a screen. But it isn't private in any legal sense. There's no confidentiality obligation. There's no privilege. And now, thanks to Heppner, there's a federal ruling confirming that.

What the court held

The Heppner decision doesn't break new ground. It applies settled privilege doctrine to a new technology. Attorney-client privilege has always required four elements:

1. A communication
2. Made in confidence
3. Between a client and an attorney (or the attorney's agent)
4. For the purpose of seeking or providing legal advice

An AI chatbot satisfies the first element and arguably the fourth. It fails on the second and third. There is no attorney. And the platform's terms of service typically disclaim any expectation of confidentiality — most reserve the right to use your inputs for model training.

Rakoff confirms what was always implicit: you cannot create a privileged communication by directing it to a machine. Privilege requires a human lawyer — admitted to practice and bound by professional obligations.

Why it matters now

The ruling matters not for its novelty but its timing. AI adoption for sensitive questions — legal, medical, financial — has outpaced the professional infrastructure to support it. People already do what the court says creates no protection.

Practical implications:

• An employee describes workplace harassment to ChatGPT, including names, dates, and their own conduct. In subsequent litigation, opposing counsel subpoenas their AI conversation history.
• A small business owner asks an AI about the legality of a contract clause they're considering. The conversation reveals strategic thinking that would normally be protected work product — but it was never shared with an attorney.
• A person going through a family dispute types out their full financial picture to "get advice." That disclosure is now a discoverable document.

These aren't hypotheticals. These conversations happen every day, by people who believe they're private.

The privilege gap

People seek legal guidance one way. The law protects those conversations another. The legal system assumes you'll call a lawyer. Increasingly, people call an AI first — or instead.

This isn't a character flaw. It's a rational response to a system that is expensive, intimidating, and slow. AI is none of those things. Privilege doctrine was designed for humans in offices, not algorithms on phones.

We call this the privilege gap: the space between what people need to discuss and what the law will protect. Heppner didn't create it. It confirmed it exists.

Closing the gap

The fix isn't to stop using AI for legal questions. That ship has sailed. The fix is to bring those conversations inside a solicitor-client relationship — where privilege applies, someone is bound to protect your confidence, and AI serves counsel rather than replacing it.

That's Blackline's approach. A licensed lawyer accepts your matter through intake and conflict check. Privilege attaches through a signed engagement. Then you get the AI workspace you wanted — supervised, protected, and defensible.

You still speak freely. You still work through your situation at your own pace. You still type instead of talk. The difference: someone stands behind those conversations — admitted to the bar and bound by professional rules.