Asking AI for Real, Honest, and Legally Accurate Answers About Filing an ERISA Disability Claim Is Asking for a Claim Denial

Welcome to Season 6, Episode 15 of Winning Isn't Easy. In this episode, we'll dive into Asking AI for Real, Honest, and Legally Accurate Answers About Filing an ERISA Disability Claim Is Asking for a Claim Denial.

What if the guidance you rely on for your disability claim isn’t just incomplete - but actively leading you toward a denial? When it comes to ERISA disability claims, generic answers, even those that sound confident and informed, can overlook critical nuances that determine whether your claim is approved or rejected. From misunderstandings about policy language to oversimplified advice about when to stop working or how to document limitations, relying on AI-generated guidance can create gaps that insurers are quick to exploit. In this episode, we unpack the risks of asking tools like ChatGPT for legal advice on ERISA claims, especially for individuals facing complex medical and vocational realities after events like a stroke. We explore why age, cognitive decline, and changing work capacity require individualized analysis, how disability carriers evaluate these factors, and why seemingly reasonable assumptions can backfire during a claim review. You’ll hear how real claimants have relied on AI for answers, only to discover that critical details were missed, misinterpreted, or oversimplified. If you’re considering filing for Short- or Long-Term Disability benefits, this episode explains why personalized legal guidance still matters - and how relying on the wrong source of information could ultimately decide the outcome of your claim.

In this episode, we'll cover the following topics:

One - The Dangerous Illusion of AI Legal Guidance

Two - When AI Gets It Wrong - and Why That Matters

Three - The Only Question That Matters - What Changed?

Whether you're a claimant, or simply seeking valuable insights into the disability claims landscape, this episode provides essential guidance to help you succeed in your journey. Don't miss it.


Listen to Our Sister Podcast:

We have a sister podcast - Winning Isn't Easy: Navigating Your Social Security Disability Claim. Give it a listen: https://wiessdpodcast.buzzsprout.com/


Resources Mentioned in This Episode:

LINK TO ROBBED OF YOUR PEACE OF MIND: https://mailchi.mp/caveylaw/ltd-robbed-of-your-piece-of-mind

LINK TO THE DISABILITY INSURANCE CLAIM SURVIVAL GUIDE FOR PROFESSIONALS: https://mailchi.mp/caveylaw/professionals-guide-to-ltd-benefits

FREE CONSULT LINK: https://caveylaw.com/contact-us/


Need Help Today?:

Need help with your Long-Term Disability or ERISA claim? Have questions? Please feel welcome to reach out to use for a FREE consultation. Just mention you listened to our podcast.

Review, like, and give us a thumbs up wherever you are listening to Winning Isn't Easy. We love to see your feedback about our podcast, and it helps us grow and improve.

Please remember that the content shared is for informational purposes only, and should not replace personalized legal advice or guidance from qualified professionals.

Transcript

Nancy Cavey [00:00:00]:
 Foreign. On AI to guide you through the ERISA disability claims process might seem efficient, but it can quietly destroy your case even before it begins. The wrong question to ChatGPT isn't just unhelpful, it can lead directly to a claims denial. Hey, I'm Nancy Cavey, national ERISA and individual disability attorney. Welcome to Winning Isn't Easy. Before we get started, I've got to give you a legal disclaimer. This podcast is not legal advice. Florida Bar association says I have to tell you that.
 
 Nancy Cavey [00:00:42]:
 And now that I've done that, nothing will ever prevent me from giving you an easy to understand overview. Disability insurance world the games the carriers play, what you need to know to get the disability benefits you deserve. So off we go. Now, I am very opinionated about this topic, so here we go. I think that asking AI for real, honest and legally accurate answers about your ERISA disability policy, filing a claim, how to write an appeal letter is a disaster. So let's say, for example, you've survived a stroke, you've returned to work, and now you're over 60 and you've begun to have cognitive and physical problems that impact your ability to work. What should you do now? Should you stop work and apply for your ERISA short and long term disability benefits? Should you consult with an experienced ERISA disability attorney? Better yet, should you just ask ChatGPT what you should do? Well, guess what? I just got a call from a prospective client asking those very questions. And the prospective clients are saying, well, we've asked ChatGPT about my disability policy or plan.
 
 Nancy Cavey [00:01:52]:
 I've asked them to speculate what a particular carrier might do in my circumstances and what their chances are of getting their disability claim approved. Now they're calling me to confirm that the ChatGPT answers are right and asking me, why should we hire you and pay you a lot of money to help me win? We could just do this with the help of ChatGPT. Really? Let's talk about what a strategic disaster this is going to be. So let's dive in. I'm going to talk about three things today. One, the dangerous illusion of AI legal guidance. Two, when AI gets it wrong and why it can matter in your case. And number three, the only question that really matters.
 
 Nancy Cavey [00:02:40]:
 What changed? Now we're going to take a quick break and while we're doing that, I want you to start chewing on those kinds of questions. When you come back, please bring a piece of paper and a pen because I'm going to drill down into the answers to these insightful questions. Got it?
 
 Speaker B [00:03:00]:
 Have you been robbed of your peace of mind by your disability insurance carrier? You owe it to yourself to get a copy of Robbed of your peace of Mind, which provides you with everything you need to know about the long term disability claims process. Request your free copy of the book@kvlaw.com today.
 
 Nancy Cavey [00:03:24]:
 Welcome back to Winning Isn't Easy the Dangerous Illusion of AI Legal Guidance. Now, filing for an ERISA disability claim, or even appealing a wrongful denial or termination is more than just asking ChatGPT questions. It's about understanding the terms of a disability policy or plan, the ERISA statute and regulations policy terms that can limit your benefits, what you have to prove to get your benefits based on your particular policy or plan, what medical proof is required, and court cases that have dealt with situations similar to yours. So let me first set the stage here. I want you to understand first and foremost that ChatGPT is not a search engine. In fact, lawyers across the United States have been sanctioned and fined by judges because they submitted legal pleadings containing cases and laws that ChatGPT made up that doesn't please the court. And when asked by a judge to produce these cases, the lawyers can't explain why these cases don't exist. They explain that they thought ChatGPT was a search engine and an accurate one, and those excuses have not impressed judges.
 
 Nancy Cavey [00:04:34]:
 Now, if you think that ChatGPT is going to give you an honest and legally accurate answer about filing your claim that will result in Your disability carrier FedExing your check within 24 hours of submitting the claim, you should think again. Why? In my opinion, relying on ChatGPT means that your claim is dead on arrival. And that's where a lawyer who practices erisa law serves as the chair of an AI legal committee and has built a closed, proprietary, trained ERISA AI product can make all the difference. And that's me. So I'm going to unpack the use of ChatGPT based on the question that I posed in the beginning of this episode and we're going to unpack this. The question was give me a real, honest, legally accurate answer. Problem with the word real. Problem with the word honest.
 
 Nancy Cavey [00:05:28]:
 Problem with the word legally accurate about how MetLife views someone going out on Ltd after age 60. That of course presumes that ChatGPT has been trained on every MetLife decision that exists about someone going out on LTD after age 60 based on ERISA standards. Now that part of the question is going to presume that the ChatGPT has been trained on MetLife views, the views of disability carriers who are reviewing claims where the claimant is over 60. And it presumes that the ChatGPT has been accurately and completely trained on ERISA statutes, standards and other rulings, Met Life claim behavior, vocational rules and the specific facts of my medical and work situation without any sugar coating. So again, this presumes that any MetLife claim behavior case law has been uploaded, that maybe MetLife practice manuals have been uploaded, that MetLife's vocational analysis tools have been uploaded, or ve reports from MetLife cases. And it presumes, unfortunately, that your medical records and your work situation information has been uploaded into the wide world of ChatGPT. So before I tell you what ChatGPT replied, let's talk about two things and I'll repeat the question because I took it apart and the question was give me a real honest, legally accurate answer about how MetLife views someone who going out on Ltd after age 60 based on ERISA standards, MetLife claims behavior, vocational rules and the specific facts of my medical and vocational situation with no sugar coating. So the two things I first want to talk about is training the ChatGPT dragon and asking questions or prompting ChatGPT to get the right answer.
 
 Nancy Cavey [00:07:38]:
 Now training the ChatGPT dragon means that you have to understand that ChatGPT is a open source, which means that you don't know what it's been trained on and you need to know that anything you input is not confidential. May I repeat that, not confidential. And I will explain later in other episodes as to why that can also be a case killer. Now this question assumes that CHAT GPD has been trained on the ERISA statute, regulations or relevant cases, court cases, your policy, the carrier's claim practices based on case law and your specific medical and vocational circumstances. And there's a saying that we learn in law school. And the saying is one should not assume facts not in evidence because that assumption will be fatal. So assuming ChatGPT has been trained properly is going to be fatal to your claim because any answer you get is not going to be factually or legally accurate. So let's Talk about the second here thing that's asking questions or prompting ChatGPT.
 
 Nancy Cavey [00:08:53]:
 Stringing questions together is not an effective prompting. You have to start with specific facts and ask precise questions, followed up by detailed follow up questions that request backup sources, sources that corroborate whatever the answer is. This is much like garbage in, garbage out, garbage, garbage, garbage. And so you need to think, why would I want to submit garbage to the ERISA disability carrier plan? Because I need These disability benefits. Okay, so in the next section of this podcast, I'm going to walk through the answer Chat B PT gave to the question Give me that real, honest, legally accurate answer about how MetLife views someone going out on Ltd after age 60 based on ERISA standards, MetLife claims, behavior, vocational rules, and the specific facts of my medical and work situation with no sugar coating. Got it. Let's take a break. Foreign.
 
 Nancy Cavey [00:10:08]:
 Welcome back to Winning isn't easy when AI gets it wrong, which is most of the time and why that will matter in your ERISA disability claim. So here's the ChatGPT answer. MetLife is more cautious when claims are filed after age 60, but they're also more likely to approve them if the medical evidence is real. Is this answer right? The answer is hell no, it's not right. MetLife is not more cautious. They're not more likely to approve a claim if you're over age 60. MetLife denies and terminates claims with people who are at 60 or even up to 64, 65 or 67. If that's the policy limits, they terminate these claims even with real medical evidence because they don't want to pay these benefits.
 
 Nancy Cavey [00:10:56]:
 Now let's talk about the real medical evidence answer and the games that disability carriers like MetLife, Unum, New York Life and others play with real medical evidence. So I want to give you an example and I want to talk about the five reasons that Unum used to justify a long term claim denial of an emergency room physician who could only work part time because of breast cancer complications, fatigue and medication side effects. Now emergency room physicians are on the front line every day. They deal with high level cognitive, emotional and physical demands. And so what does a disability carrier like Unum do when the policyholder is diagnosed with stage four metastatic breast cancer? Now you would think that they would do the right thing and pay benefits, but you know that would be too easy. And often Unum and other disability carriers or plans will wrongfully deny a claim and force the claimant, including physicians or lawyers, into into stressful and needless litigation. And that's what happened in the case of Kman versus Unum Life. It's a case out of the state of Washington and Dr.
 
 Nancy Cavey [00:11:58]:
 Kirsman was an emergency room physician at Harborview Medical Center. She was diagnosed with stage four metastatic breast cancer. There's a 27% survival rate. After five years, the cancer had metastasized to her liver and ribs. She took a full year of medical leave and underwent chemotherapy, a lumpectomy, hormone treatment and and a liver ablation. A year later, with the approval of her oncologist, she returned to work in the emergency room on a 50% schedule with no night shifts. So Unum in their infinite generosity, paid a year of residual disability benefits based on her reduced hours and then terminated her benefits, alleging she could work full time. So UNAM and properly determined that an emergency room physician's job is light duty and improperly suggested that the treating oncologist created reasons to restrict Dr.
 
 Nancy Cavey [00:12:51]:
 Kirsman to part time work. After the denial, notwithstanding the fact that Unum had already accepted the part time work restrictions and paid residual benefits, they improperly argued it exercising, doing yoga or vacationing with their family were inconsistent with complaints of fatigue and they improperly ignored the known side effects of her treatment. They improperly ignored the obvious immological issues with working in an emergency room setting. Now fortunately, a federal judge threw out Unum's bogus denial. But this case illustrates that even in the view of real medical evidence and in a compelling case, Both Unum and ChatGPT got it wrong. Now the ChatGPT answer to the question goes on to say here's the truth. Like really? Medlife knows that people over 60 often don't want to leave work unless they have to. And most long term disability claims filed after age 60 are legitimate because people want to reach retirement age.
 
 Nancy Cavey [00:13:55]:
 People want to protect their Social Security. People often try to push through impairments. People nearing retirement rarely manufacture disability claims and MetLife has known that for decades based on claims data. Well, is any of that true? Well, most people who are not disabled do want to work till retirement age and they do want to protect their Social Security retirement benefits and the increase in the amount of those benefits that they can get by virtue of continuing to work and contribute to the Social Security system. People who are disabled might also want to continue to work to increase their Social Security disability or retirement benefits. And so those statements are maybe true. But how does that make a claim, an LTD claim after age 60, more legitimate than a claim made by a 50 year old? What does the word legitimate even mean in this context? Most claims are legitimate. I wouldn't represent somebody who isn't legitimate.
 
 Nancy Cavey [00:14:58]:
 And disability carriers still won't pay benefits and we have to fight about it. What about the rest of this answer? Not everybody tries to push through impairments, whatever that means. Impairment is not the standard of disability. Disability policies or plans generally require that you be unable to perform the material and substantial duties of your own occupation, as those terms are defined in each policy. Or plan. There is no uniform definition of disability in a policy. In fact in policies. And there are many times that I'll see 10 or 15 versions of Unum policies and there can be multiple versions of policies for just the same employer or classifications of employees.
 
 Nancy Cavey [00:15:45]:
 So there is no uniform policy that defines disability. Now is it absolutely true that people nearing retirement rarely manufacture their disability? Well, disability carriers or plans think the exact opposite. And they often suggest that the policyholder simply doesn't want to work until retirement and they manufacture a claim. So I would say to you that 95% of that answer is wrong. And if you relied on that wrong answer in formulating your strategy, filing your claim, writing an appeal, what do you think is going to happen? Nothing good. So what are the right questions based on the prompts? Ask those wrong questions and you get the wrong answer or an answer that doesn't matter. Now. Wrong question number one is why are long term disability claims made by someone older than 60 legitimate? Who cares? So as I've said, what does the word legitimate mean? Does age make a claim more legitimate? And I would have been asking ChatGPT, well, when I wouldn't have been using ChatGPT, but I would have asked the follow up question that says what's the basis or what's the definition of legitimate? In your answer, why does age make a claim more legitimate? Cite the sources that you are relying on and provide me with copies of any case law or studies that disability carriers use in reaching the opinion that older claimants claims are more legitimate.
 
 Nancy Cavey [00:17:20]:
 And by the way, how does this relate to my particular claim since I've given you facts about my medical history, my vocational history, and you've got the policy. So there's a lot of questions that should have been asked in follow up that were not asked. Wrong question number two, why do people want to protect Social Security? Well, what does the word protect mean? Does my policy include a contribution for Social Security benefits in my policy? Now that might be relevant, but then the next question is going to be, well, how does that help my claim? And by the way, please produce for me all the sources that you are using in reaching this conclusion and including a copy of the particular documents or studies that you're relying on. Wrong question number three, why do people over 60 push through impairment? And again, the question here is, well, what is impairment? Is that required by my policy? What's the definition of disability in my policy? What sources support your conclusion that people over 60 push through impairment? And more importantly, why? You know, what are the sources that would support the answer in this case, in my particular unique situation, that they would conclude that I, me, myself and I would push through my impairment, whatever that word means. Wrong question number four. Why do people nearing retirement age manufacture disability? Well, again, we, we don't know what the word manufacturer means. We don't know what the word disability is. We don't know what the source is.
 
 Nancy Cavey [00:19:04]:
 We don't know what documents they're relying on. And they meaning chatgpt. And how on earth is this relevant to this particular case based on the facts that were just put out in the open ChatGPT world? Wrong question number five. What is MetLife's claim data being referenced? Produce it. I want to see it. And how does it impact my particular claim? Now, the answers are going to change when these questions are asked. And more importantly, I will tell you, if you ask those questions multiple times, you will get different answers. More importantly, none of these are the right questions.
 
 Nancy Cavey [00:19:44]:
 So what's the next wrong ChatGPT answer? MetLife internally acknowledges that high level professionals often hold on until age 60. This strengthens your claim and not weakens it. Well, there's a lot of presumptions here. What is the basis of the conclusion that MetLife internally acknowledges that high level professionals. Hold on. What does hold on mean? What's the impact of age 60? How does it strengthen your claim? Produce case law or data or MetLife manuals that cooperate and support this conclusion. Now, another question would be why does MetLife care that you push through? I mean, to me that's a really stupid question. The real question is what changed? Why is it that you've been able to work and now all of a sudden you can't work? Because that's exactly where MetLife is going to focus on.
 
 Nancy Cavey [00:20:44]:
 What changed medically? What changed vocationally? Why is it that this person can no longer do their own occupation? Have they had an improvement in their condition? What was the improvement? Has there been a decline? What was the decline? How is that decline relevant to their ability to do the material and substantial duties of their occupation? Have they been compliant with treatment? Have they followed a regular treatment schedule by an appropriate physician, a specialist? What does the doctor have to say about the person's change in their physical exam findings or their complaints or their ability to function? Those. I mean, that's the right question, but you can hear. I posed a heck of a lot of questions that would be relevant in looking at a person's policy, medical records, vocational history, and trying to formulate a winning application or potentially appeal a wrongful denial or termination. Did ChatGPT do just that? No, I did it based on my years of experience as an ERISA disability attorney. Got it. Let's take a break.
 
 Speaker B [00:21:54]:
 Are you a professional with questions about your individual disability policy? You need the Disability Insurance Claim Survival Guide for Professionals. This book gives you a comprehensive understanding understanding of your disability policy with tips and to dos that will assist you in submitting a winning disability application. This is one you don't want to miss. For the next 24 hours, we are giving away free copies of the Disability Insurance Claim Survival Guide for Professionals. Order yours today@disabilityclaimsforprofessionals.com.
 
 Nancy Cavey [00:22:33]:
 Welcome back to Winning Isn't Easy. The only question that matters is what changed Now? We started this podcast telling the story of a person who survived a stroke, returned to work, and now over age 60, has cognitive and physical problems and is thinking about applying for their disability benefits. Now does MetLife care that you push through? If you were receiving residual benefits, they weren't happy because they're paying you based on the difference between what your pre disability earnings are and your post disability earnings on a monthly basis. Either way, you're now claiming total disability and the question is always going to be what changed? And that question is going to happen when you have been working but you have the residuals of a problem and you haven't applied for your disability benefits. The question becomes what changed? And it doesn't matter if you've been collecting residual benefits or not. They're always going to ask, well you worked. Or say, look, you worked yesterday, you didn't work today, what changed? And you better have an answer to that and one that has been well developed prior to you pulling the plug and saying you're done. So the right questions that they're going to ask and that you should be considering and please don't ask ChatGPT that will make or break your claim is what's changed medically and generally.
 
 Nancy Cavey [00:24:01]:
 They're going to look in the at the last three years, has there been a progression of the symptoms? What are those symptoms? What's the objective documentation of the progression of the disease and the symptoms? What do your medical records say? Do your medical records talk about the location of your symptoms, the nature of those symptoms, the duration of the symptoms, the intensity of the symptoms, the frequency of the symptoms, and an example of how each symptom impacts your ability to do your occupation? And of course you need to understand what occupation is defined in yes, because it could be your occupation is performed by your employer for your employer in the local economy, the national economy, the Dictionary of Occupational Titles. You need to make sure that your medical records are addressing the definition of your occupation as you develop the symptoms and functionality in your medical records. Now, of course, the ultimate issue is going to be how does this impact your ability to perform your own occupation? And why are you stopping now? What is the magic here? Have you just decided you're done? Have you decided you're going to retire? You're not getting along with your boss, You've decided that you're just tired of pushing through and what you think is going to matter to the disability carrier plan? So there's got to be an explanation in this strategy as to why you're stopping now. And ultimately we need to understand, is your doctor going to support your restrictions and limitations? And as part of that, you and your doctor need to understand the definition of occupation. Occupation is not a dictionary definition. It is a definition based on the terms of your policy or plan. It could be you have to be unable to do the material and substantial duties of your occupation, or some of them, or the primary ones, or the most important ones, and, and the word occupation might be the occupation, as I've said, that you're doing for your employer in the local economy, the national economy, the Dictionary of Occupational Titles. That's going to require on your part an analysis of what duties you can't perform based on your symptoms and what duties you're having problems with.
 
 Nancy Cavey [00:26:26]:
 So you want to be educating your doctor before you pull the plug and say I'm done. But you also want to consider what's in that personnel file. Are you being written up for problems or are you still the stellar employee and getting 10 plus reviews? What would your coworkers say if asked about the difficulties that you might have? Have you gotten a vocational evaluation by a vocational friendly, a claimant friendly vocational evaluator that will connect the dots between your occupational duties, the definition of disability as defined. Both terms are defined in your policy or plan and connect the dots between your symptoms and your doctor's opinion about your restrictions and limitations. You've got to explain why, from a vocational and medical standpoint, you can't do your occupational duties. And then I think this requires a lot of planning. When do you plan on stop working? Is that consistent with what's in your medical records? Because if it isn't, you got some work to do before you really should stop work and apply for benefits. Now, if you're at risk for termination because you're getting bad reviews, you may want to speed up the process like now, but that again requires a lot of planning in terms of strategy.
 
 Nancy Cavey [00:27:47]:
 This is what it is we need to do. This is how we need to do it. This is what needs to be in the medical records. And, and by the way, this is the date that works for me. Because the other factor here is what happens to the benefits that you have through your employer. You might have life insurance, hard to replace. You might have pension. What's going to happen to your pension and pension contributions? Are you going to be getting bonuses? Should you be hanging on to get those bonuses because not only do you want the money, but it will impact potentially the calculation of your benefits? What happens to your group insurance benefits? Are you going to lose them? Do you have to cobra them? Can you move your coverage to your spouse's policy? And you need to also understand Social Security disability requirements.
 
 Nancy Cavey [00:28:35]:
 Why many disability policies or plans have offsets or reductions for the receipt of Social Security benefits. And you got to apply because if you don't, they're going to act as if you're getting Social Security and they will reduce your long term disability benefits by not only the Social Security you get, but the Social Security benefits your dependents get. So you've got to figure out here what your net benefits are after offsets. That was a laundry list. Did that ChatGPT answer consider all of those factors? Heck no. But would an experienced ERISA disability attorney consider that in assisting you in deciding whether you're going to apply, when you're going to apply, what evidence you need? Of course. So I'm going to ask you this question now. What do you really think about those ChatGPT questions and those answers? Do you really, really, really think you would get your short term disability or long term disability benefits? ChatGPT will rarely give you the right answer because it doesn't know the right question to ask based on your unique circumstances and it doesn't know how to ask the right follow up questions and it doesn't provide documentation for the basis of their answers.
 
 Nancy Cavey [00:29:53]:
 In my view, only an experienced ERISA lawyer can listen to your story, review your policy or plan, analyze your medical records and answer your questions. Better yet, in addition, they'll ask you questions that you never thought to ask that you need to consider in making this decision as to if and when, and they will help develop a winning strategy to get the disability benefits you deserve. Of course. Are you going to have to pay for it? The answer is yes, you're going to have to pay for it. But isn't that better knowing you have an excellent chance of getting your benefits than throwing caution to the wind being a cheap person? And I don't mean to insult you by asking ChatGPT the right questions that you think are going to get the right answer. Well, the answer is heck no. This wraps up today's episode of Winning Isn't Easy. Thank you for tuning in.
 
 Nancy Cavey [00:30:51]:
 If you found this episode helpful, take a moment to like our page, leave a review, share it with your family and friends, and don't forget to subscribe to this podcast. I would love to hear from you, particularly about this episode, and at the beginning of each episode description you'll find a link where you can send us your questions or comments. I would love to feature your questions about the use of AI and my thoughts about that process. Now join us next week for another insightful discussion. There's going to be more about AI and in particular how using AI before you retain an attorney can also doom your claim. Got it. Thanks for listening.