Regular readers of this blog will know that I took a hiatus from legal marketing after seeing the tactics we used to use on behalf of clients lose their "oomph." Many still worked, but not to the degree they once had, and in many cases, not enough to justify the kind of retainers we were charging.
Since getting back into the game with our Associate product and taking calls with former clients to learn how their marketing has evolved, one thing has become clear. Law firms are embracing AI generated content, otherwise known as synthetic content.
Google knows which content is AI generated
The trend is troubling because it runs afoul of bar association rules as well as Google guidelines, and in the case of Google, they 100% know which content is AI generated and which isn't. How do they know? I suggest reading this Nature paper on SynthID-Text to better understand where things are headed with AI content watermarking.
In short, frontier AI models will tag your content as AI generated with an invisible watermarking scheme. The tag gets added if you edit in Claude and copy paste into WordPress or another CMS.
For more, read this primer on the Anthropic site.
Winning content is unique, not synthetic
I wrote in a previous post that the only way to win with written legal content in the age of AI is to leverage unique fact patterns, war stories, and client case studies. Get in the weeds and demonstrate expertise, tell us something we don't know and can't easily look up.
Google search has a crawl budget. The last thing they want is to waste resources inhaling some dry, over optimized article reciting a basic point of black letter law.
But it's more than that. Claude and other AI tools now tag content with watermarks so it's clear which content was generated by AI and which wasn't.
Transparency is now the rule
The policy change is driven by the EU AI Act, which is concerned with transparency in AI. In the past, some law firms would hire copywriters to generate blog content and publish it on the firm website under a partner's name. The behind the scenes process of generating, editing, and ultimately publishing the content was a black box. No longer. Each piece of AI content the firm generates will have the watermark signature tagged to it via a watermark detection API.
So what is the best approach?
Real editing and real writing augmented by AI.
The use cases for AI in a law firm's content process are numerous. Our AI agent we built to revolutionize legal marketing, called Associate, will update page titles, meta descriptions, build and fix links, and yes, write some content, but it's a matter of degree.
There is no issue with augmenting a great piece of content using AI to, for example, fix outdated passages that don't reflect changes in a statute. Using an AI agent to make these updates is a no brainer.
But that's also not what I am seeing with the firms I am talking to. It seems many have inverted the process by using AI as the primary writer and adding superficial edits after the fact.
Your AI content is visible
Law firms need to understand that their AI content is visible. An entire blog written using AI is now visible via the SynthID-Text that was tested on 20 million Gemini responses.
If your firm is investing in AI generated content, it's crucial to understand that the degree to which your site leverages AI will now be visible.
Want this running on your firm’s site? Request access →