Using generative AI in your appraisal preparations

Generative Artificial Intelligence (AI) refers to a class of AI systems designed to create new content, such as text, images, music, or video, by learning patterns from existing data. A subset of generative AI is Large Language Models (LLMs), which are specifically trained to understand and generate human language and perform tasks such as text generation, translation, and summarisation. LLMs are, for the most part, where GPs might choose to use AI with their appraisal preparation. That’s what we’ve focused on here.

To make sure you’re using these tools responsibly, effectively and safely in your appraisal preparations, AI literacy is really important. You need to understand the limitations of AI and ensure you’re checking all the outputs are accurate. AI can help you prepare for your appraisal. It should not do the work of reflecting for you.

Your Responsible Officer (RO) may have rules about the use of AI in appraisal, and we recommend that you do not use AI to write reflections on learning. You could use it to provide prompts for your own reflection but, by avoiding overuse of AI for reflection, you can make sure you maintain the tools you need to be a reflective GP.

It's also worth noting that AI tools can often produce a lot of suggested text, and summarising that yourself could help to ensure you’re checking the information being provided and properly reflecting on it.

Make sure you let your appraiser know whether you have used AI in the preparation of your evidence.

Confidentiality issues can be a concern when using AI. Always check that you’re using AI in a way that is consistent with any relevant data protection or employment requirements across all your medical roles. In particular, patient identifiable information should not be entered into a web-hosted generative AI platform.

Any use of AI as part of your appraisal preparation should be made clear to your appraiser. Be clear and transparent about how you’ve used AI, including what model and software you’ve used. Your appraiser can then assure themselves that you’ve used it safely and are following any local rules or guidelines around the use of AI.

The GMC doesn’t yet have guidance specifically for using AI in appraisal. However, they do state that they expect doctors use their professional judgment and apply the principles outlined in Good medical practice when using AI tools in relation to their medical practice.

We would always recommend being open about how and when you’ve used AI in preparing your appraisal portfolio, and to speak to your RO, or their appraisal team, in advance of your next appraisal if you have any questions or concerns. Including some reflection, either written or verbally in the appraisal discussion, on how you used AI in your practice in the past year is a good way of ensuring you’re assessing its impact on your practice and patient safety. 

AI prompts are specific instructions or questions given to an AI model to generate a desired response. There are many tutorials on how to structure prompts online with examples.

Some AI models allow you to upload reference material which can help mitigate the risk of AI ‘hallucination’. Generative AI hallucination (or confabulation) occurs when an AI tool generates information that is incorrect or nonsensical, despite it sounding plausible. The AI might fabricate details or fill in gaps with invented information, presenting it as if it were true. This happens because generative AI relies on patterns and associations learned from its training data and when faced with incomplete, ambiguous, or unfamiliar inputs, the AI may produce responses that seem coherent but are actually false.

AI can be used to gather and summarise data and information from many different sources. For example, you might want to include a summary of an online journal you read to reflect on for a learning log or your appraisal portfolio, keeping in mind the GMC emphasis on quality and not quantity of supporting information. Generative AI can create outputs and answers, but it does not 'understand' these outputs. As such, while generative AI can be used as a reference and information synthesis tool, it must be used in a critical way, understanding its limitations and for users to be prepared to challenge its findings.