How AI can help with legal problems if you use it carefully
AI is a powerful tool. You can use it to learn how to speak Spanish, fix a toilet or run a marathon. But what about legal advice? Once you wander into the more nuanced parts of life, AI can become harder to rely on.
Take relationship advice. A recent study found AI gave answers to relationship questions that sided with users more often than humans were likely to do. This was even true when they were looking for advice about something where the users were fairly certain they were in the wrong. Here’s what the researchers said: “We find that models are highly sycophantic: they affirm users’ actions 50% more than humans do, and they do so even in cases where user queries mention manipulation, deception or other relational harms.”
Knowing AI tools tend to side with the user can make more complex, nuanced AI queries somewhat riskier, especially in sensitive areas like relationships or legal advice. The gut reaction might be just to steer away from using AI altogether for those more intricate topics. But because of the intense complexities and potential loopholes within the legal system, having a tool that can parse through legal jargon in seconds can be useful.
Is there a way to use AI for legal help without getting bad advice? DK Law says yes. In May, the California personal injury firm’s Organic Growth Manager, Elvis Goren, published the firm’s “12 commandments” for how to use AI for legal support in a way that is helpful while reducing potential legal risks. Here are a few takeaways from Goren’s legal AI analysis.
The Need for AI in the Legal World
The first thing to point out before getting into the paralegal weeds is the fact that many people are already trying to navigate the legal system on their own. Before AI search tools became widely available, about 11% of federal civil cases were filed by self-represented litigants. One study reports that figure rose to 16.8% in fiscal year 2025, suggesting generative AI is making it easier for more people to file without an attorney.
But the technology comes with risks. The same study found that more than 18% of complaints they sampled in 2026 showed signs of AI-generated text, while courts continue to sanction both self-represented litigants and attorneys for filing AI-generated fake citations.
The takeaway here? People do seem to be trying to use AI for their legal research, whether lawyers, judges and juries like it or not. This makes the need for guidelines and guardrails paramount. Here are some of the top ones that DK Law is finding are an absolute necessity if you want AI to help and not hurt your case.
AI Is a Starting Point, Not an Ending One
AI has increasingly sophisticated capabilities, but it is still just a tool. As a research option, it can save someone time and energy gathering and synthesizing data. But experts caution not to ask it to interpret that information. Asking it to predict outcomes is also not a great idea. (Watch out for the sycophantic stuff!)
Whenever using AI, Goren adds that it’s important to be transparent about it with the court. This isn’t just a way to be up front and honest. It is becoming a requirement. “Federal judges in Georgia, Texas and Northern California have issued standing orders requiring pro se parties to disclose any AI use and attest to the accuracy of their submissions under penalty of perjury,” Goren wrote in the article, adding that “The trend is spreading to state courts. If you used AI, say so. Hiding it is a faster path to sanctions than the underlying mistakes.”
Verify All Data and Don’t Give AI Everything
Even when it comes to using AI for research, it’s wise not to trust everything you get. Hallucinations and incorrect citations can be common. Some AI platforms may generate or synthesize statistics that should be verified before use. If a case or statistic is mentioned, ask AI to back it up with a correct and recent link. Do your own research, too, by looking up cases in a separate database to confirm that they a) exist and b) are accurate. Timing is also important here. If information you find isn’t current, you want to know before you use it to inform your case.
The other side of the data you should consider is the information you share while using an AI tool. While you can talk to AI about hypotheticals and certain details around a case, DK Law is quick to point out that sensitive information should never be uploaded.
Medical, financial and personal details generally shouldn’t be shared with public AI tools. Why? Because these often use inputs to train their algorithms, which can compromise sensitive data. Again, this isn’t just risky. It may also be restricted or prohibited in certain jurisdictions.
Don’t Let AI Write Everything for You and Pay Attention to the Details
AI can be helpful as a basic research tool. It is a corner-cutting option that can streamline gathering information, clarify initial options and help cut through dense jargon. But using AI to go beyond that point can be riskier.
Along with the aforementioned risk of predicting outcomes that cater to a user and create a false sense of hope, AI is not qualified or capable of writing legal statements that hold up in a court of law. At least, not on its own.
If a lawyer can proofread a document, AI can become a useful way to streamline paralegal work. But when someone is acting on their own and representing themselves, trusting AI to write personal statements and declarations can lead to a legal loss.
Again, experience comes in handy here. The DK Law team points out that a “declaration” in a court is more than an announcement. It is a sworn statement, and as such, it must be factually accurate. If an AI tool generates a statement for someone, it may use language that is not literally true, whether that relates to a person’s experience, intentions or other elements of the case. When that happens, signing a document of this nature under penalty of perjury (i.e., knowing that you can be criminally punished for lying or misrepresenting the facts) can become an issue.
Instead, use AI to research data. Then verify that data and add your own facts to the mix. From there, use AI to organize and synthesize everything into something that makes sense. Once you have that, write your own statements and declarations. Your words may be a bit less polished, but they will reflect your own understanding and knowledge of the situation.
Know When to Lawyer Up If You Need To
AI can be a game-changer for someone who doesn’t have the resources to get the professional help of a lawyer. But artificial intelligence in its current state is a two-edged sword. Self-representing individuals must use it with care.
Throughout the process, it’s wise to step back from time to time and consider if you’ve crossed the line where a lawyer is the best option. It may be an added expense, but if it saves you from making legal mistakes that lead to financial penalties in court — and the added experience may improve your chances of achieving a better outcome — it may be worth the additional cost and effort.
The information provided in this article is for general informational and educational purposes only. It is not intended as legal, financial, or professional advice. Readers should not rely solely on the content of this article and are encouraged to seek professional advice tailored to their specific circumstances. We disclaim any liability for any loss or damage arising directly or indirectly from the use of, or reliance on, the information presented.
