ChatGPT Without the Liability
The privilege-safe AI workflow for paralegals, HR managers, and SMB owners. 8 lessons grounded in ABA Opinion 512, the Heppner ruling, and the $110K Oregon sanction.
Here’s the quiet problem: about 1 in 20 employees has already pasted confidential business data into consumer ChatGPT this year. At the average company, it happens hundreds of times per week. Some of that data was source code. Some of it was client names and financials. Some of it was regulated health or HR data. And ninety percent of the people who did it don’t realize they’ve done anything wrong.
In February 2026, a Southern District of New York judge reportedly held that AI chat logs are not protected by attorney-client privilege. In April 2026, a federal judge in Oregon imposed a $110,000 sanction against two attorneys whose AI-generated brief contained 15 fabricated case citations. The ABA’s Formal Opinion 512 — published July 2024 — already made clear that a supervising attorney is liable when a paralegal pastes client data into a non-BAA-covered AI tool.
This course teaches the framework that keeps you out of that pattern. Eight lessons, two hours, free first two lessons, certificate at the end.
What You'll Learn
- Explain attorney-client privilege, work-product doctrine, and why AI chats now sit outside both
- Identify the 5 categories of content that must never enter consumer AI — with profession-specific examples
- Differentiate Consumer, Team, Enterprise, and specialized AI tiers and their real data posture
- Apply a Kovel-adjacent AI workflow at attorney or principal direction with the right paper trail
- Audit and clean your Claude Memory (and ChatGPT Memory, and Gemini Memory) in under 10 minutes
- Draft a one-page AI use policy your team will actually follow
After This Course, You Can
What You'll Build
Course Syllabus
Who Is This For?
- Paralegals at small and mid-size law firms
- HR managers and HR business partners
- Small business owners and solo consultants
- Financial advisors and tax preparers
- Anyone on a regulated-data team who uses ChatGPT or Claude at work
Frequently Asked Questions
Who is this course for?
US-based paralegals, HR managers, SMB owners, consultants, and financial advisors — any professional who handles privileged, regulated, or third-party confidential data. It's explicitly NOT for licensed attorneys (they have ABA Opinion 512 and need firm-level rollouts, not this course) or clinicians (who should take OpenEvidence for Bedside Nurses instead).
Is this legal advice?
No. This course teaches a workflow informed by ABA Formal Opinion 512, *U.S. v. Heppner* (SDNY Feb 2026), *Couvrette v. Wisnovsky* (D. Oregon Apr 2026, $110K sanction), and published firm memos from Jones Walker, Morgan Lewis, K&L Gates. For privilege-specific decisions about your own matter, consult a licensed attorney.
What's the difference between this and 'AI for Paralegals'?
This course teaches the framework — how to use ANY consumer or enterprise AI safely across ANY profession that handles sensitive data. 'AI for Paralegals' is a deeper playbook specifically for paralegal workflows (citation verification, Kovel engagement in depth, enterprise legal AI tools). Take this one first.
Will I get a certificate?
Yes — a verifiable CPL-XXXXXX certificate after completing the 8 lessons and capstone audit. Add it to your LinkedIn or compliance portfolio.
What's updated in the April 2026 version?
Cites the *Couvrette v. Wisnovsky* ruling (largest AI sanction to date), the *U.S. v. Heppner* third-party disclosure analysis, and the post-Jan-2026 ChatGPT for Healthcare vs consumer tier distinction. Material is framework-first, news-second — it'll still be accurate when the specific rulings fade.
