Directory submissions. They come around every year, and each time you swear you’ll do better. Be more organised, spend more time collating cases and referees in advance. But at the heart of the matter, legal professionals quite simply do not have the time to prepare effective submissions without some serious help.
So what if there was another way? A way to streamline the directory process and improve the impact of the content you’re generating. Enter the elephant in the room, AI.
Now, this might be met with gasps of horror but AI really can be a useful tool to help even the most experienced professional with directory submissions.
Is anyone really thinking about AI for directory submissions seriously?
Let’s put the cards firmly out on the table… we think that when it comes to directory submissions, AI can be a game changer.
Did you know that the majority of top 20 firms we talk to are already utilising AI (for example Copilot) to edit and expand on directories profiles and matter overviews? And as well as asking us for help with directories strategy, we’re seeing more and more wanting our help to implement effective AI into their directories process, including writing prompts for them that work? Makes you think… doesn’t it?
So, how can we use AI for directory submissions?
Writing for clarity and results.
Perhaps the first thing to springs to mind, when considering using AI for directories, is the way you could get it to help you summarise things – whether that’s overviews, profiles, matter descriptions – and break down the complex legal jargon into accessible summaries for researchers.
You can ask it to comment on the gaps in your content and things you can do better. To suggest ways you could stand out more or adhere to the category or directory guidelines. AI is a great consultant, when it comes to things like this. And, although the content it spits out might need some work by a human, the basic information will certainly be there, if used correctly.
Get help with the data.
Another aspect that can be simplified with the use of AI is generating and organising your referee spreadsheet. It does this by analysing the raw data – such as names, contact information, and case associations. It can then organise referees into a coherent list, flag missing or incomplete information (such as email addresses), and link referees to the most appropriate cases for them to comment on. When it comes to formatting the referee spreadsheet, AI-powered templates can automate the alignment of data fields to match directory-specific formatting requirements. By using AI, your spreadsheet can be populated in seconds, saving time and ensuring accuracy.
Highlight confidential information.
Ironically, as we will come on to below, the tools that everyone fears because of confidentiality and plagiarism risks, can help you identify confidential information such as client names, and case numbers and flag sections of text containing these details and other potentially privileged information such as case outcomes. AI can replace sensitive names or details with placeholders throughout the submission, and also help to rewrite descriptions to maintain clarity while removing any identifiable information. It can also tailor the content by placing it in the appropriate section of the form, such as Publishable and Non-Publishable matters. By integrating AI into the confidentiality review process, legal professionals can enhance accuracy, reduce manual oversight, and confidently submit well-protected and compliant applications.
Free up billable time with automation.
And last, but certainly not least, AI frees up billable time. By automating repetitive tasks like drafting, manually inputting referees and formatting, lawyers can spend more time (doing their jobs) but also reviewing the submission afterwards.
The risks can be high with AI, though, if used incorrectly.
While AI can be an amazing tool to use, it gets its information from the internet as well as from you. We all know the importance of fact-checking. And that AI should never be the sole author of your submission. Human oversight is essential to ensure accuracy and compliance. But there are some bigger risks that need careful consideration before you jump on the AI bandwagon.
Get prompting right.
AI works best when being prompted by someone knowledgeable about the subject matter. One of the pitfalls many fall into when first using AI, is bad prompting. If you ask AI to write you something with a vague, non-specific prompt, you’re going to end up with a generic answer. The aim is to be concise and plain with language so that you can get across key information to the researcher quickly, without them having to shuffle through the technical details to understand the significance of the case you’re presenting.
With that in mind, when writing your prompt, you need to consider your audience, the data that the AI will need and provide clear instructions, structure and most importantly, examples such as past submissions to give the AI that all important context. You can set boundaries for AI responses, to keep them focused and factual, and include a command to never make assumptions, if there is any incomplete information. Some AI tools can provide citations and sources, but there are also third party verification plug-ins available to flag any inconsistencies or misinformation. In short, the more detailed and precise the instructions, the less likely the AI is to hallucinate and the closer it’ll be to what you want.
Confidentiality is paramount.
Confidentiality is probably the biggest concern everyone has when doing directory submissions – and when using AI, actually. Let’s be clear – there is absolutely no way that any directory submission information (publishable or otherwise) should be used in a non-secure system, however tempting.
Ultimately, each firm is different in its use of AI and the measures it has put in place – and these need to be adhered to. These could range from using trusted AI platforms from reputable providers with strong data security policies, such as encryption, data anonymisation, and compliance with regulations like GDPR or CCPA, and avoiding public-facing tools like free versions, through to only using secure versions of these tools (such as team or non-learning settings) or even only using on-premises AI installations or those hosted on a secure, private cloud.
Now, while you can use AI to anonymise and flag confidential information, the real fool proof method is to ensure this is done before inputting the data. In this way you would ensure that nothing confidential would go into any system… but that might not really work, in terms of making the best of the opportunities it presents.
It takes a while to learn.
The other cons, while not as serious as confidentiality, are still worth mentioning. There is a learning curve (for both you and the AI). It takes time to learn how to use AI tools effectively, and equally, it takes some time for the AI in question to learn your style and tone of voice in order to replicate it. There is also the possibility that AI may misinterpret complex legal details to start with, so you can’t rush at this, as a solution.
Is AI really worth the hassle then?
Exploring AI responsibly, to streamline processes and improve submissions should, we believe, be encouraged – so long as you understand that while AI is powerful, it is not infallible… and neither are you, when using it. Treat its outputs as drafts that require refinement rather than finished products. Implement strict checks, provide clear risk-assessed guidance and policies for use, and rely on human oversight at all stages to safeguard against hallucinations and manage confidentiality risks.
The key to success when using AI is understanding its capabilities and limitations. We believe that by embracing AI as a collaborative tool rather than a competitor, we can combine efficiency with human ingenuity, improving both our personal and professional lives.
It can seem daunting to try something new when tried and tested methods work just fine, but if you want to give it a go we recommend initially just having a play around with the various tools available to you. Try it out with some low level tasks… you don’t need to jump straight in!
And if you want nothing to do with directories (AI or not), working with an agency like ELE could be the smartest move your firm or chambers makes.
Take a look at our directories hub here, which has all of our directories resources, all in one place. Or, if you want to have a chat about how we might be able to help your firm or chambers really level up your submissions, get in touch here.
