Exploring Gen AI Opportunities for Plain-Language Writing

I’ve always prided myself on being a good writer. I spent years honing my grammar skills and natural instinct for sentence flow and ‘what just sounds right’. However, it has also become second nature to rely on some writing tools in my day-to-day work. Like most, I utilize spell check. I also rely upon tools like Hemingway App to ensure my plain-language projects are at the appropriate grade level for the intended audience. Yet I haven’t incorporated Generative AI into my writing toolkit.

What is holding me back from potentially a more efficient writing process? Even if I believe my skills are above par, could Gen AI help me improve my writing, particularly for plain-language projects?

The Law Society of Saskatchewan, my home jurisdiction, in the Guidelines for the Use of Generative Artificial Intelligence in the Practice of Law, encourages lawyers to engage in continuous learning about AI and its implications for legal practice.

There are also numerous examples of opportunities for using Gen AI to improve our practices. The Generative AI Playbook from the Law Society of Alberta provides several examples, including document preparation:

• Creating agendas, memos and contracts;
• Drafting correspondence, proposals and outlines;
• Identifying risks and inconsistencies in contracts; and
• Creating presentations.

With respect to the creation of public legal information, Gen AI has been identified as a potential tool for delivering public legal information and education. See for example the 2022-2023 Bridging the Gaps Legal Needs Assessment from the Alberta Law Foundation.

Recently, the Centre for Public Legal Education Alberta (CPLEA) built upon this report and explored how Gen AI could:

• help provide translated public legal information in more languages and formats;
• support self-guided help for users; and
• present legal content in multimedia formats to better meet diverse needs.

Key findings from the CPLEA team include:

• AI supports human expertise but cannot replace it.
• AI translation tools show potential for short-term content but there are review hurdles.
• Plain language editing and writing AI tools improved internal workflows.
• AI video production is still premature, but shows promise with voice, captions/transcripts, and editing.
• Chatbot and interactive web tools are not viable for CPLEA – yet.

CPLEA has also developed initial guiding principles to support safe, effective, and ethical use of Gen AI across their work. These guiding principles generally are in line with the Gen AI guidance provided by law societies. Examples from the law society guidelines are included, with a focus on the Law Society of Saskatchewan Guidelines identified above. The CPLEA Principles include:

• Always have a human in the loop. Gen AI can’t replace the need for human review, oversight and final decision-making.
  • Saskatchewan Example: Duty to Supervise and Delegation:
    Generative AI tools should be treated as equivalent to nonlawyer assistance, and their outputs should be reviewed for accuracy and conformity with the lawyer’s professional obligations.
• Consider use and disclosure. For example, CPLEA doesn’t use Gen AI to generate content from scratch or for legal research. When Gen AI is used, the tool and how it was used will be disclosed.

• • Saskatchewan Example: Duties of Competence and Diligence:
    Overreliance on AI tools is inconsistent with the ethical and responsible practice of law. A lawyer must apply their independent and trained judgment when acting for clients.
  • Saskatchewan Example: Communication
    A lawyer should consider disclosing to a client if they intend to use generative AI in carrying out their representation.
• Consider issues such as privacy, bias, security risks, trust and safety, and the role of AI tools.
  • Saskatchewan Example: Duty of Confidentiality
    Consideration must be given to whether client confidentiality or privilege may be lost as a result of using a particular AI tool.
• Only use tools that will ethically increase efficiency.

See the article, Artificial Intelligence in Public Legal Education: An experiment from the bottom up, by Judy Feng for more in-depth information on the findings of the project and guiding principles. See also a recent LinkedIn post by Amy Salyzyn for a list of guidance documents available.

With the CPLEA findings and principles in mind and with my law society’s guidance, I plan to explore various Gen AI tools to improve my plain writing skills and efficiency. In particular, I will explore how well these tools can assist in writing for particular audiences, with a focus on generational differences. See my post, Good Communications: Don’t Forget Generational Considerations, for more information on this topic. Stay tuned for my report. I also invite those who have already begun this exploration to share their experiences.