Jonathan Bick in the New York Law Journal: The Most Significant Ethical Risks of AI in Legal Practice

Date Published: 8/20/2026
Originally published in the New York Law Journal
As artificial intelligence becomes deeply integrated into everyday legal workflows, courts and bar associations continue to reinforce a foundational rule: attorneys remain strictly accountable for the accuracy, confidentiality, and ethical propriety of AI-assisted work product.
In a recent analysis published in the New York Law Journal, Jonathan Bick examines the intersection of generative AI and the Rules of Professional Conduct, breaking down the five most critical ethical hazards practitioners face:
Breaching Client Confidentiality (Model Rule 1.6): Entering sensitive, non-public client information into public or open-source platforms that retain data for model training.
Failing to Verify AI Work Product (Model Rules 1.1 & 3.3): Submitting AI-generated research, briefs, or citations without independent verification, particularly when systems produce “hallucinated” legal authorities.
Surrendering Independent Judgment (Model Rules 5.1 & 5.3): Relying on AI as an autonomous decision-maker rather than treating it as a supervised tool.
Improper Billing Practices (Model Rule 1.5): Charging traditional hourly rates for work drastically expedited or completed by AI without adjusting fees to reflect actual labor.
Omitting Informed Client Consent (Model Rules 1.4 & 1.6): Using AI tools that handle client data or materially shape litigation strategy without upfront transparency and consent.
The piece also reviews less common, high-risk misuses—including evidence manipulation, improper witness coaching, bias discrimination, and deceptive litigation tactics—while detailing proactive steps firms should take to discuss and memorialize AI use at the outset of representation.
Read the full article in the New York Law Journal
Related Practices: Patent, Intellectual Property & IT
Related Attorney: Jonathan D. Bick
Related Industry: Technology





