AI Risks for Illinois Lawyers to Watch in 2026
AI risks for Illinois lawyers rarely announce themselves before they cause damage. During a presentation to the Chicago Bar Association, the results were revealing. A basic chatbot readily drafted routine correspondence. But it also fabricated background information, generated stereotyped images, and issued a confident verdict from a two-paragraph dispute.
For lawyers carrying professional liability coverage, these aren't abstract concerns. A fabricated case citation or a biased AI output that reaches a filing becomes a malpractice exposure. Understanding how these failures happen, not just that they're possible, helps firms build better safeguards.
Three risk categories stood out:
Hallucinated content
Built-in bias
Deepfake authenticity problems
Together, they point to the ownership standard that defines Illinois' approach to AI in legal practice.
AI Risks for Illinois Lawyers Start with Hallucinations
Hallucinations remain the most immediate AI risk for Illinois lawyers, and a Chicago Bar Association presentation demonstrated the problem in two ways.
The first was familiar. Generative AI tools can cite cases, statutes, and facts that don't exist. This is a distinct risk category, separate from ordinary drafting errors.
The second example was less expected. When asked to research the judge's own background, ChatGPT produced a summary describing him as a longtime Carbondale attorney focused on business and insurance litigation. The tool then correctly noted that he'd since become a judge. But it didn't stop there. Unprompted, it offered to look into unverified rumors and controversies involving judges in the surrounding counties.
That offer illustrates a risk beyond simple hallucination. AI tools will readily speculate and present their speculations with the same confident tone as verified facts. A lawyer using AI for background research, due diligence, or opposing counsel research could easily receive fabricated leads dressed up as findings.
The presentation's message stayed consistent throughout. Fabricated citations and fabricated facts fall into the same category of risk. Both require the same remedy: independent verification before anything reaches a filing or a client. Skipping that step turns a convenient shortcut into a costly mistake.
Built-In Bias Compounds AI Risks for Illinois Lawyers
Bias is baked into many AI tools. When prompted for images of a successful attorney, the AI system returned older white men in suits, surrounded by law books and cash. A prompt for a surgeon produced similarly narrow imagery, while a prompt for a thug generated stereotyped images of Black men.
These outputs reflect patterns in the training data used to train AI systems, not deliberate programming choices. But the effect is the same regardless of intent. A firm using AI-generated imagery for marketing, training materials, or client-facing content risks reproducing that bias without realizing it.
The presentation also featured a live demonstration of AI Judge, a website that generates instant verdicts from user-submitted dispute descriptions. In the example shown, a homeowner and a neighbor each described a dispute over a shed fire started by a child playing with a lighter. The tool issued a confident, detailed verdict awarding damages, citing legal principles that had little connection to the facts presented.
The verdict sounded authoritative. It also reflected no actual legal analysis, no jurisdiction-specific law, and no verification of either party's claims. Holistically, that gap between tone and substance is exactly what makes AI-generated legal content risky once it reaches a courtroom or a client relying on it for guidance.
Deepfakes and the Professional Conduct Ownership Standard
Authenticity concerns extend beyond text into video, audio, and images. Deepfake content often consists of a person's voice, face, or body being digitally altered so they appear to say or do something they never did. Moreover, he illustrated the point with a side-by-side comparison showing how convincingly a face can be swapped onto unrelated footage.
For courts, that capability threatens the reliability of video and audio evidence. Litigation involving recorded statements, security footage, or expert testimony now carries an added authentication burden.
Illinois addressed these risks with a single accountability standard rather than tool-specific bans. The Illinois Supreme Court Policy on Artificial Intelligence took effect January 1, 2025. It permits the use of AI without disclosure, provided the resulting work remains grounded in fact and law. The standard is put simply, "You sign it. You own it."
That standard connects directly to a professional conduct rule requiring lawyers to stay current on the benefits and risks of relevant technology to maintain competence. A lawyer doesn't need to become an AI expert. But using a tool without understanding its failure modes falls short of that competence standard.
Managing AI Risks for Illinois Lawyers in 2026
Rather than disappearing, AI risks for Illinois lawyers will shift as technology matures. The presentation made clear that hallucinations, bias, and deepfakes already show up in ordinary use, not just edge cases. Each risk traces back to the same root cause of AI tools presenting output with confidence that outpaces their actual reliability.
Conclusively, Illinois' policy response puts the responsibility on the person using the tool. Lawyers don't need to disclose AI use, but they do need to verify it. That means checking citations against primary sources and reviewing AI-generated images for bias before they reach a client or the public. It also means treating any audio or video evidence with appropriate scrutiny.
Overall, the presentation closed with a nod to a broader shift already underway. Chief Justice John Roberts has observed that young lawyers and partners alike will need to stay adaptable as this technology reshapes legal practice. That adaptability, paired with careful verification, separates responsible AI use from a malpractice exposure waiting to happen.
For firms building risk management practices around AI, these three categories- hallucinations, bias, and deepfakes- offer a useful starting checklist. ISBA Mutual works with Illinois lawyers to strengthen exactly this kind of AI risk management. For help with professional liability and cyber coverage, contact the team at ISBA Mutual Insurance Company.
