Artificial intelligence (AI) is rapidly transforming the legal profession—but for mortgage default law firms, where accuracy, efficiency, and compliance are paramount, the adoption of AI has been met with caution. Many firms are intrigued by the promise of AI but hesitant to embrace it due to concerns about confidentiality, ethics, and control.
Can AI really help a mortgage default law firm? What exactly does it do, who can benefit, and how do you manage the risks associated with it? These are all questions that I have struggled with as a law firm owner and as someone who is much more comfortable making legal arguments than attempting to learn something as technical as AI.’
While I am not the most tech-savvy attorney, I do understand that AI is here to stay, and if I don’t understand it and implement it where I am able, I will be left behind. So, I reached out to Eyal Leeder, an expert in AI who has created an AI platform that was tailored specifically to assist the mortgage default industry. Eyal answered my questions in a language that I could understand. I left our conversations feeling excited to get started, and I thought the conversation might be helpful to other attorneys trying not to be left behind. Below are some of the questions that Eyal helped me with.
I see AI everywhere—in articles, webinars, in the news, at conferences and in casual conversations with friends—but I don’t understand how a law firm can benefit from AI.
Eyal: AI provides you with access to tools that can dramatically reduce the time that a person would have to spend on repetitive tasks, things like populating forms, creating policies, template generation, searching internal documents for answers, ensuring billing is submitted properly, marketing, and creating draft newsletters. The tasks it can handle are really endless.
If it can do all of this, then why do you think there is such hesitation to move forward with implementing it in mortgage default law firms?
Eyal: Mortgage default firms handle high volumes of sensitive data—from borrower financials and loan documents to court filings and foreclosure notices. As a result, firms are understandably cautious about adopting any new technology that might jeopardize client confidentiality, introduce legal or ethical risks if outputs are inaccurate, or disrupt compliance workflows tied to investor, agency, or court guidelines. In addition, a default firm doesn’t really manage one workflow, but dozens of separate flows based on client-specific requirements and guidelines. Introducing brand new technology into such a complex environment can certainly appear daunting.
And then, of course, there’s a fear of hallucinations and a general lack of clarity about which tools are safe, which are not, and how AI fits into legal practice.
Hallucinations?
Eyal: Yes, hallucinations. It is when AI generates information that may sound correct but is factually wrong. Think of it like this: a five-year-old might come to you and excitedly tell a story about meeting space aliens at school. It is clearly not true, but to hear the kid tell it, with all the excitement and hand waving and sound effects, it’s a very convincing presentation. It is also very likely that the kid completely believes it, too, even though it only happened in their imagination.
You see, large language models such as ChatGPT, Google’s Gemini, xAI’s Grok, and others all generate their responses in a probabilistic manner. They have been fed vast amounts of information and trained on it, but what they actually utilize is a vast set of probabilities. For example, the model knows that if the question relates to basketball, it is likely that the word “Michael” would be followed by the word “Jordan”, but it doesn’t actually understand who Michael Jordan is or what he accomplished. Every word, every sentence, and every paragraph is created using probability, not certainty or prewritten information. So, a model can easily respond with something that it calculates as being PROBABLE but didn’t actually happen. That is what we would call a hallucination.
What are some of the most practical, day-to-day uses for AI in mortgage default law firms?
Eyal: It can start at the beginning and assist in accepting referrals and collecting information from referral documents to open the file. It can draft pleadings policies, communication and training materials, search documents for specific information, generate templates, and review invoices for accuracy. It can also run QC checks on every file on a regular basis and flag exceptions.
In my platform, AskTuring, the firm can upload all of their documents, like client guidelines and requirements and directives, fee agreements, policies, regulatory materials, etc., into the secure, private index, and the firm employees can then “talk to the documents”. For example, the employee can ask the platform what a certain client allows in billing for a specific task, and the Index will generate the answer based on its review of all of those documents, and also provide the user with a link to the specific page and paragraph where that info was pulled from.
AI isn’t a replacement for lawyers or paralegals—it’s a powerful assistant that can handle routine, repetitive, or data-heavy tasks at speed and scale. Some of the most beneficial tasks AI can handle for a law firm are:
- Document Automation—Generating first drafts of complaints, motions for default, affidavits, and notices, and populating forms using data from the firm’s case-management system (CMS).
- File Review & Quality Control—It can flag missing or inconsistent documents in foreclosure or bankruptcy files. It can be used to verify if assignments or loss mitigation documents are in place, or if files have missing or incorrect fee approvals.
- Data Extraction & Summarization—AI can read and summarize loan documents, notes, deeds, or payment histories and pull key data points (loan #, default date, borrower info, legal description, etc.) from large PDFs.
- Bankruptcy & Litigation Support—Monitor PACER or court dockets for bankruptcy filings, draft motions for relief from stay or proofs of claim with less manual effort, and compile payment histories needed for POC.
- Compliance Checks—It can ensure that file timelines meet GSE, CFPB, or investor guidelines, and can automate file audit reports for clients or servicers.
- Billing and Invoicing—Read and extract information from incoming invoices, create outgoing invoices, and ensure the correct backup documents are attached based on state, investor, and client requirements.
One of the biggest concerns that I have with using AI in the mortgage default industry is security and protecting private information. Is it even realistic to think that we can use AI in light of the security requirements of our clients?
Eyal: Security is a big issue in our industry. The AI tool you use should have SOC2 certification and not share data with the foundational models. Self-hosted AI or enterprise-level privacy controls are big factors in the level of security.
Is AI going to replace the workforce?
Eyal: No, we are always going to need humans to think and analyze. Especially in our industry. We have to keep humans in the loop to maintain legal integrity. Anything generated by AI needs to be reviewed for accuracy by a human. Besides, I have learned that no matter how long you have been in the industry, there is always something that makes you go, “That’s the first time I’ve ever seen this happen!” Those are instances where the AI requires human guidance and decision-making.
Of course, with foreclosure volumes expected to continue to rise, law firms have to improve efficiency. The rising workload and complexity of the work make manual processes unsustainable. By having AI complete or even start some of these tasks, law firms will save a significant amount of time each week, which frees them up to focus on tasks that truly require human analysis or interaction.
Final Thoughts
I am so appreciative of the time that Eyal spent answering my questions and explaining how AI could make a real difference in the productivity of my firm. I walked away from the conversation with a completely different understanding. I actually am excited about moving forward with implementing AI in my law firm. I still have some concerns, so I will approach it as Eyal suggested, in baby steps, in order to learn as we go and understand the potential of AI to improve our processes and run the practice more efficiently.
From our conversation, I learned that AI offers real, measurable benefits to mortgage default law firms—from improved efficiency and accuracy to better compliance and client service. Yes, there are risks—but they can be managed. With the right tools, training, and policies in place, AI can become a strategic asset rather than a liability.
