Key Takeaways:
- AI is ethical to use in PI when governed by existing legal guidelines
- You cannot delegate legal judgment or responsibility to AI
- Understanding how your AI tools work is part of your duty of competence
- Confidentiality applies to AI tools just like any third-party vendor
- You must review and verify all AI output before using it
- Efficiency gains improve your practice but shouldn’t inflate billing
—
Despite the rapid rise of AI across the legal industry, many personal injury firms are still holding back. A 2025 report by 8am found that 59% of PI firms have not yet adopted any legal-specific AI tools, and 37% point to ethical concerns as a major barrier. This raises the question: Is it ethical to use AI to prepare and draft settlement demands?
In 2024, the American Bar Association (ABA) issued its first formal usage framework on generative AI ethics for lawyers, known as ABA Formal Opinion 512, alongside the Model Rules of Professional Conduct. At a high level, the opinion asserts that the same duties governing your practice today still apply; AI simply changes the context in which those duties are carried out.
If you’ve been waiting for the profession to catch up before you commit, consider this your green light with a few guardrails attached.
Here’s what you need to understand as a PI attorney before your next demand goes out the door.
Is It Ethical for PI Lawyers to Use AI? What the ABA Says
As long as you use AI within the same professional standards that already guide your work, using AI in a PI practice is ethical. But when you’re questioning yourself, ABA Formal Opinion 512 is a helpful guideline to reference.
Its core message is that generative AI is merely a tool, meaning you cannot outsource your professional judgment to it, especially in something as consequential as a settlement demand. The Model Rules of Professional Conduct also establish that certain duties are non-delegable — so no matter how sophisticated the technology, the responsibility stays with you.
- Rule 1.1 (Competence): The duty of competence has always required staying current with technology, and right now, generative AI is the most relevant technology in the room. According to Rule 1.1, you must understand the tools you’re using to practice law. That means knowing how the AI generates output, where it can fail, and what it can and cannot see.
- Rule 1.3 (Diligence): Under this rule and the broader framework of attorney oversight, you are responsible for reviewing what the AI produces before it goes to an adjuster, opposing counsel, or a court. Every AI-generated demand your firm sends is your demand with your signature and your client’s case on the line.
From an ethical standpoint, AI doesn’t reduce the attorney’s role in advocacy; it increases the importance of oversight.
Where AI Use Becomes Unethical
The ethical risks of AI are tied to how it’s used. In the context of settlement demands, failure points include:
- Blind reliance on output: Violates competence and judgment
- Failure to verify facts or citations: Risks misrepresentation
- Inputting confidential data into unsecured tools: Confidentiality breach
- Delegating legal judgment to AI: Improper reliance
- Inflating time or fees despite increased efficiency: Unreasonable billing
These aren’t new categories of misconduct, but you should still be aware of how they present in a new context.
We’ve also seen inconsistency emerge as a significant ethical issue in traditional demand workflows. Large-value cases often receive exhaustive attention, while smaller cases may not get the same depth of review simply because firms lack the time to manually surface every category of pain and suffering, emotional impact, or future damages.
AI can help close that gap by making more comprehensive demand review economically feasible across all case sizes, rather than just the largest ones.
Can You Upload Client Files to an AI Tool Without Violating Confidentiality?
When it comes to confidentiality and the work-product doctrine, carefully evaluate whether your AI tools provide adequate protection.
Uploading case files (medical records, police reports, billing statements) to build a demand in an unsecured or public AI platform exposes protected client information to a third-party system. Under Rule 1.6, your duty of confidentiality requires that you understand how a platform handles your data before uploading anything.
- Does it store your inputs?
- Does it train future models on your case materials?
- Is it HIPAA-compliant and SOC 2-certified?
The work-product doctrine protects your litigation strategy from disclosure to adverse parties. But if your AI vendor lacks adequate security controls, you may inadvertently waive privilege or expose client information.
AI platforms purpose-built for legal practice can help you prevent this ethics violation with documented security standards and clear data policies. Closed-loop legal AI systems keep sensitive case data within a controlled, isolated environment, preventing exposure to broader public AI ecosystems.
Generic consumer tools like ChatGPT were not designed for attorney-client confidentiality, and the ABA has been direct that using them with client data requires serious scrutiny.
Is Client Disclosure Required When Using AI for Demand Drafting?
Under Rule 1.4, you owe clients reasonably complete information about their matters. In some cases, that includes disclosing your use of AI if it impacts how services are delivered or how their data is handled. In PI, where clients are often vulnerable, anxious about their cases, and trusting you with deeply personal medical information, err on the side of transparency.
It’s not necessary to disclose each tool used over the course of representation. However, client disclosure is likely required when:
- AI plays a significant role in producing the final demand
- It’s relevant to a client’s representation decisions
- AI use requires sharing confidential data with a third party
A brief sentence in your representation agreement, such as “Our firm uses AI-powered tools to assist in preparing demand packages; all work product is reviewed and approved by a licensed attorney,” covers the disclosure obligation and builds trust.
The Non-Delegable Duties of PI Practice
The most ethical use of AI in a PI practice is to accelerate the parts of the process that don’t require lawyer thinking. These are tasks where AI, used correctly, retains quality while adding consistency and speed.
Examples of these tasks within the demand process include:
- Summarizing medical records
- Organizing chronological treatment timelines
- Identifying patterns and gaps within records
- Structuring the narrative framework
- Drafting initial versions of routine sections
However, case valuation judgment remains firmly within the category of non-delegable duties. Strategic decisions like these are still legal work requiring attorney judgment:
- Which injuries to emphasize
- How to frame pain and suffering for a particular adjuster
- How aggressively to position the settlement value of the case
- What tone and negotiation strategy will be most effective in the demand
The AI doesn’t know your client’s face, the county’s reputation on juries, or that the insurance company always fights soft tissue claims hard. You do, and the Model Rules require that you keep doing it.
One of the biggest long-term risks with AI is the erosion of domain knowledge. We’ve seen how easy it can be for users to start rubber-stamping outputs rather than actively evaluating the facts and reasoning behind them. Ethical AI use requires staying engaged with the substance of the case, not just approving the final draft. Simply having a human in the loop isn’t enough if the review becomes passive approval.
The real safeguard is active engagement with the facts, reasoning, and strategic framing behind the demand.
How to Use AI Ethically in PI Demand Drafting
For PI attorneys, demand letters are the foundation of a case’s value, strategy, and credibility, which makes them one of the highest-stakes areas to get right from an ethical standpoint.
Here’s how you can put ethical, practical AI to use in PI demand drafting:
- Choose purpose-built platforms. AI tools trained on personal injury practice, HIPAA- and SOC 2-compliance, and that pull only from your uploaded case documents carry far less ethical risk than general-purpose AI. When the AI’s training aligns with PI standards, the output is more accurate, and your review burden is lower.
- Establish firm policies. Under the Model Rules, managing attorneys are responsible for ensuring that subordinates, including non-lawyer staff, use AI appropriately. Be sure to put your AI use policy in writing and train your team on what it covers.
- Review everything before it goes out. Legal AI isn’t meant to remove attorneys from the demand process — only to give them better tools to evaluate cases thoroughly. Assume all output needs work, and put your name on it only when you’d be comfortable defending it to the client or a disciplinary board. In our view, the goal shouldn’t be an AI-assisted demand but a verifiably true one. Attorneys should be able to quickly trace claims back to supporting records, treatment timelines, and source documentation before a demand ever leaves the office.
- Document your review. If a question ever arises about how a demand was prepared, you want a record that an attorney reviewed and approved it. Good practice management means creating that paper trail as part of the workflow.
Ethics as a Competitive Advantage
Roughly 80% of legal professionals expect AI to have a transformational impact on their jobs within five years. In personal injury law, that transformation is already underway.
The firms that come out ahead won’t be the ones blindly automating everything or refusing to adopt AI altogether. They’ll be the ones that understand where human judgment still matters, where AI improves consistency, and how to build workflows that keep attorneys actively engaged in the reasoning process.
At the same time, the ethical standards around AI are still evolving and likely will continue evolving quickly as the technology advances. Firms that stay adaptable, informed, and actively engaged with how these tools are used will be better positioned than those treating AI adoption as an all-or-nothing decision.
In our view, that balance, not full automation, is where ethical AI adoption in personal injury law is ultimately heading.
Want to See What Ethically Built AI Looks Like in Practice?
AI Demand Pro is purpose-built for personal injury attorneys. It’s HIPAA- and SOC-2 compliant, trained on PI practice standards, and designed to keep every output grounded in your case documents with attorney oversight baked directly into the workflow.
Unlike general-purpose AI platforms, AI Demand Pro operates within a secure closed-loop system designed for sensitive legal data. As the gold standard for handling sensitive information, our closed-loop system keeps case data within a controlled environment, preventing it from flowing into open consumer AI ecosystems. This results in stronger privacy protections, more defensible outputs, and a workflow centered around verification, traceability, and attorney review.
Book a Demo to See Attorney-Grade AI in Action
FAQ
Does using AI increase my ethical risk as an attorney?
When AI is applied within the existing framework of competence, confidentiality, and supervision, it’s no more risky than delegating work to a human.
Is it unethical to rely on AI for parts of legal work?
Using AI to assist is acceptable; replacing your independent judgment with it is not.
Can using AI improve ethical compliance?
Yes. When used correctly, AI can reduce human error, improve consistency, and help ensure important details aren’t overlooked, which strengthens the overall quality of your work.
Do ethical obligations change depending on the AI tool I use?
Your obligations don’t change, but your risk exposure does. Tools with stronger security, better data controls, and more relevant training reduce the likelihood of ethical issues.
What are the ethical considerations when using AI for demand letters?
When using AI, attorneys must ensure the demand reflects the record, avoids exaggeration or mischaracterization, and aligns with their legal strategy. You’re responsible for the story the demand tells.
