Misinformation about AI and client confidentiality persists in the legal profession, but the key question is whether lawyers know how to properly vet and govern the technology they use.

In Morgan v. V2X, Judge Braswell offers a thoughtful, practical take on AI use in litigation—reminding lawyers (and even pro se litigants like Morgan) that when it comes to confidential data, it’s less about the tool itself and more about how responsibly you handle what you put into it.

Early court decisions confirm that using AI does not automatically waive privilege—but failing to ensure confidentiality, proper settings, and attorney direction just might.

Using AI with confidential client information is not inherently riskier than the cloud tools lawyers already trust—provided it’s done with the same level of diligence, configuration, and vendor scrutiny.

“As AI tools become affordable, effective, and widely adopted, a lawyer’s failure to use them may soon be viewed not as caution—but as a breach of the duty of competence.”

Deploying an autonomous AI in your practice isn’t using a tool—it’s appointing an agent, and you’re liable for what it does within its authority.

The real disruption isn’t that AI replaces lawyers—it’s that it destabilizes per-seat software pricing, automates commoditized legal work, and gives clients new leverage to challenge how legal fees are structured.

Even the perfect prompt cannot rescue bad context engineering—because what your AI sees, structures, and prioritizes behind the scenes determines your output at least as much as the question you type.

AI has made BigLaw faster and more profitable—but not cheaper. Instead of expanding access to justice, efficiency gains are flowing upward, leaving legal costs high and the justice gap intact.