Generative AI is a tool that can create new content, such as text, images, or music, based on some input or data. It uses deep learning models that can learn from large amounts of data and generate outputs that are relevant, coherent, and sometimes creative.
As an accountancy firm, we are always looking for ways to enhance our services, boost efficiency, and stay ahead of the curve. We could see the potential of generative AI to help with tasks like report writing, data analysis, crafting presentations, and generating insights. While the technology is still evolving we wanted to ensure that our teams were on the journey as soon as possible, making each adoption step smaller.
After evaluating various options, we chose Microsoft 365 Copilot. This choice was influenced by its robust data protection measures, ensuring our data remains within our Microsoft 365 environment and is not used to train external models. Additionally, Copilot integrates seamlessly with the Microsoft 365 applications our team already uses, providing a familiar interface.
Our journey so far
1. We created a policy for all colleagues that explained the risks of generative AI, including ethics, trust, hallucinations, and the importance of checking the work before releasing. To maintain control and security, we restricted the use of AI tools to Microsoft 365 Copilot exclusively.
2. During a regular company-wide meeting, our CEO introduced the policy and emphasised that generative AI is a valuable new resource meant to enhance—not replace—colleagues’ skills, urging everyone to explore its benefits in accordance with the policy.
3. Our team created and conducted firm-wide training on generative AI, highlighting its support features instead of job risk. The training programme covered:
the basics of generative AI
how to prompt Copilot effectively (the way to ask questions of AI)
its ethical use; and
practical, interactive, activities relevant to areas like marketing, education, project management, tech, and accounting.
We began by addressing employees' worries and concluded by ensuring their concerns were addressed.
4. Three months after the training courses, we surveyed our colleagues to evaluate their usage. The outcomes were highly positive, suggesting that our policy and training were successful. The training received high praise with a 4.55 out of 5 rating for its effectiveness and practicality, which increased our colleagues' confidence in using generative AI. Over 75% of our people are now applying AI occasionally in their tasks, mainly for drafting documents, creating content and conducting research. Many are using it more often. Benefits noted include enhanced quality, efficiency and creativity in their work.
Looking ahead
We are very proud of the progress we have made so far in integrating generative AI into our business. But these are only the first steps on our AI Adoption journey.
Short term
We will organise training sessions focusing on the two key areas where generative AI can add the most value:
creating and drafting content; and
research and information gathering.
The sessions will provide tips and examples based on the examples our colleagues told us about in our survey.
We are also creating a forum to encourage colleagues to share their successes and failures using Microsoft 365 Copilot with peers and the project team.
We are also reviewing the other forms of AI to see how they could fit into our business processes and are actively discussing with our software suppliers to understand how they will bring AI into their software. We want to ensure that we are always up to date and ahead of the curve when it comes to using the latest and most innovative AI tools.
Integrating AI comprehensively into a professional practice is challenging. Most professionals are open to technology but require evidence of its benefits before altering their routines. Future software will likely include built-in AI, removing the need for deliberate adoption. Currently, users must figure out how to incorporate AI into their existing workload. Therefore, our strategy is to create a safe space that reduces risks, promotes the use of AI, and facilitates the sharing of inventive AI uses within the team. Projections indicate that Microsoft 365 Copilot adoption rates are expected to reach 35% by the end of 2025, reflecting significant growth in enterprise AI integration.
Reproduced from Legal Abacus, the magazine from the ILFM, with the permission of the publishers.
