Last month, a new research paper came out. Access to justice in the age of AI. The authors ran 4.6 million civil cases in the US federal courts and found that since the appearance of ChatGPT, the number of pro se cases and a pro se case is when a lawsuit is being filed without a lawyer. So, the number of pro se cases has increased sharply. Specifically, in categories like consumer credit, civil rights, and employment discrimination. It had been a steady 11% for 20 years straight and is now 17%. Yes, that 6% increase is quite sharp because the number did not change for 20 years. In absolute numbers, the pro se lawsuits rose from 23 to 41,000 per year. Now, what's interesting here is that despite the rise in pro se cases, the volume of cases with lawyers stays flat. Listen to this one more time. The number of cases with legal representation, which is about 190,000 per year, did not drop. It stays exactly the same. What this means is that there is zero evidence that AI is currently capable of replacing lawyers or lawsuits filed with lawyers. Instead, it creates a new type of demand. The core change in this example is that people who previously would not have filed at all are now filing. It is a horizontal expansion of the legal business. And the reason I'm saying that it's a horizontal expansion is because it doesn't change and doesn't disrupt the existing business. Clients are not saving any money because they weren't paying anyway. The audience for these cases are not the people who would otherwise hire a lawyer and are trying to save money. It's the people who wouldn't have filed before. The only party that feels the pressure from this type of newly created business is the judge. Judge is the only party in this equation that feels the consequence of this expansion in a negative way is because there are many more cases now, and activity in courts has increased by 158%, but the number of judges and overall court capacity hasn't changed in 40 years. And these pro se cases become an additional workload. And this workload will keep growing. To sum it up, the disruption that AI is doing in the legal field comes in the form of making something that used to be expensive or difficult much more accessible. And that's great. But this disruption happens in an area where we weren't paying anyway. So, if there is a startup that comes up with a way to reduce the cost by an order of magnitude, that becomes a disruption. But when there is a new market opportunity, it is not really a disruption of the market. It is the expansion. If disruption were to happen in the legal field, the first party that would feel it would be corporations that use big law firms. As you can see, AI does not automatically disrupt a market just because the technology exists. Your customer still has to find you and still has to understand what your product does differently. But ironically, while your product or business does not have to be AI-native, your customers will most likely find you through AI search. And if you're building a business or managing a company, you need to know whether your brand shows up in those answers, how it is described, and which competitor is being recommended instead of you. And this is where Ahrefs Agent A comes in. Agent A is an AI marketing assistant built on top of your marketing data. Agent A has a variety of skills that can automate any marketing job on your list. You can audit how AI tools describe your brand, track how your company gets mentioned, and analyze brand sentiment. Scan products of your competitors, find content gaps, generate keyword ideas, and turn all of that info into marketing reports. Ahrefs has spent the last 14 years indexing every page, every link, every keyword, every ranking, every AI mention. And Agent A can manipulate this data, combine it, cross-reference it, and act on it in real time. It is trained on every Ahrefs tool, metric, and use case to make the most of your SEO data, web analytics, insights, brand radar, rank tracker, site audit, and more. It also connects with tools like Notion, Slack, HubSpot, WordPress, and Linear. So, the work doesn't live inside one dashboard. It can actually move into the tools your team already uses. So, if you're building a product and want to understand not just your SEO, but your visibility in the new AI-driven discovery layer, check out Agent A from Ahrefs using the link below. And now, back to the video.

So, let's go through every single one of these and see what the market tells us. AI to audit law firm bills. Now, what does this mean? I'm going to give you a scenario. Let's say a corporation hires a law firm and the firm completes the work on a certain legal matter that they were hired for. And now the firm sends the bill to the corporation. The corporation looks at it and goes, "Okay, this charge is valid, this charge is valid, but the time you spent in a taxi going from point A to point B, I'm not paying for. And also, you put this charge twice by mistake." Now, this is a gross simplification of this process. My point is there could be duplicate entries, block billing, non-compliant charges, and a million other things in those invoices. And that's what the industry is built around. Now, in reality, I think the best way to understand whether disruption is really happening is to look at the market of the legal ops professionals. Part of the jobs that legal ops do is reviewing those invoices and doing audits. Of course, doing those audits is not the only thing that they do on a regular basis, but it is a meaningful part job. And in reality, there is no disruption in this field. Automation is not happening. Legal ops head count is flat to slightly growing. And lastly, before AI auditing is enabled for a corporation, that corporation has to agree to use AI auditing for their needs. This is much harder than it seems because being someone who builds those tools for large corporations. You get to see how wary people are when it comes to the use of AI.

The second area is automating research-heavy work that would be typically done by junior associates. Now, this area is most visibly disrupted from the tech standpoint. Harvey AI, Thomson Reuters CoCounsel, LexisNexis, Legor Cloud released a version of the model specifically for legal research. And you would think that as research or document drafting get faster, the time spent on a matter will be shorter because well, there's less to do. In reality, law firms became more profitable than ever because of AI. Am Law 100 firms posted record revenue in 2025. Average billing rates jumped 9% in the first half of 2025 alone. Profit per lawyer is up 53% since 2019. Am Law 100's standard rates, once again, standard rates, cracked $1,000 an hour for the first time in history in 2025.

And the third area, insourcing. Now, the core problem is that the work that in-house teams are doing with AI, contract drafting, certain legal research, compliance docs, is the work they were already doing internally. Legal AI apps make legal teams faster at their own jobs. It's the same thing when you give cloud code to a developer to speed them up. It speeds them up, but it doesn't replace what they're doing. The high complexity, high stakes work, litigation, M&A, regulatory investigations, still goes to the outside counsel. It is not a question of purchase access to this AI tool and start doing that you couldn't do before. When companies outsource, it is almost always for highly specialized and niche work. And nobody is going to insource an M&A transaction or an intellectual property matter to a generalist in-house general counsel. Or make them stand up in court or in front of regulators to carry that risk because they used AI.

So, to sum it up, at the bottom of the market, AI enables people to file lawsuits that they wouldn't be able to afford if they had to hire a lawyer. But at the top, it makes expensive lawyers faster and more productive. In both places, the pattern is the same. AI expands the system, but it doesn't take the money out of it. The value that AI adds, the value that you can put a dollar sign on, is not disruptive. It exists and it'll grow, but it is 100% not a replacement of lawyer work. It makes something far more accessible, something that wasn't accessible before, but it does so in areas where we weren't paying anyway. As always, we hope this was helpful and we'll see you in the next episode. Bye.