Video Summary

This video analyses a recent Employment Tribunal judgment in Scotland involving Sandy Peggy and Dr. Upton, focusing on claims related to the use of changing facilities by trans women. The speaker expresses concern that the tribunal may have fabricated quotes or misrepresented information from previous cases, notably from the For Women Scotland and Maya Forstater rulings.

The discussion highlights issues in the tribunal's ability to accurately quote other judgments and warns about the potential dangers of relying on AI for legal interpretations, noting how it can lead to erroneous conclusions. The video underscores the importance of validating legal references and critiques the tribunal for selectively quoting to create misleading narratives.

The speaker assesses specific paragraphs of the tribunal's judgment, pointing out inconsistencies and errors, and concludes that several parts of the judgment are fundamentally flawed and warrant an appeal, suggesting a lack of proper adherence to established legal principles. Overall, the content serves as a critical examination of the judgment's validity and the challenges posed by integrating AI in legal contexts.

Transcript

This is really quite extraordinary. I now strongly suspect that the Employment Tribunal for Scotland has either elaborated on its judgment with AI, such as ChatGPT, or is just making things up. This is evident from parts of the judgment that are quoted but just don’t exist.

In this video, which I don’t often like to do, I will talk about the same topic in two videos on the same day. I prefer to keep it varied, but there is so much going on that I need to draw your attention to these additional points. I’m obviously referencing my 'Black Belt Secrets' video, which I’ll link below because it works perfectly in tandem with this. There’s a whole load of content in that video for you to watch as well, and I won’t touch on the same things too much that I’ve talked about there. These are new points, and I’m absolutely blown away.

So, let me get to it. This is the Employment Tribunal Scotland judgment for Sandy Peggy, Dr. Upton, and NHS F Health Board. Long story short, Sandy Peggy and Dr. Upton worked at NHS F. Dr. Upton is a trans woman, a biological man identifying as a woman, using the female changing rooms. They got into an argument, and a report was made to NHS F, after which Sandy Peggy was suspended. She then brought a claim under the Equality Act for various issues, including harassment and discrimination.

Now, the ruling of the tribunal is a bit strange. Sandy Peggy succeeded in her claim against NHS F but didn’t succeed on some other points. I’m not going to dwell on those, but the focal point of this judgment seems to be regarding the use of changing facilities by trans women. It appears that instead of providing reliable information, the tribunal has cherry-picked facts and fabricated quotes that simply don’t exist, firstly from For Women Scotland, and also from the Maya Forstater case and another case. I will look at some of those with you in this video.

I’m going to borrow these posts from someone on X because they outline some of the points neatly, which is helpful to present on-screen. As a lawyer, whenever I see a quote or a case reference, I always verify it. I almost never take anything at face value; if I see a statistic or a number, I go and find the original source to see where it came from. If you don’t do that, you’re not doing your job correctly.

The big danger with AI is that it can generate 'hallucinations', misquote, paraphrase, and manipulate information to fit a desired conclusion. Unless you specifically instruct AI not to fabricate or twist the information, you can end up with gobbledygook, giving erroneous conclusions and quotes that don’t exist. Lawyers have gotten into serious trouble, even contempt of court, for including case references in their documents that are fictitious. It’s an embarrassingly stupid mistake that should never happen.

On that note, I’m developing an AI legal assistant, 'Black Belt Barrister Pro', which I am training not to make these errors. It’s not a replacement for legal advice but a tool to help you with your legal research problems. You can find that linked in the description below.

Now, Anna Palmer notes here that a tribunal can reconsider its judgment to correct simple errors, known as the slip rule. They can return and say, "I didn’t mean that; I meant this.” This is often referenced in criminal cases where there is confusion over days and years, for example. However, Anna states, and I agree, that if the quotes attributed to the Employment Appeals Tribunal in Forstater and the Supreme Court in Lee and Ashers were actually from other cases, it's easy enough to correct the reference. But in this case, it doesn’t seem to be the case.

I’ve looked at some of these and they are bizarre. The quote mentioned in the court article is from the EAT judgment in Forstater. How do we know it isn’t in that judgment? We can search the judgment on BAILII, which is a free database where you can find these judgments yourself.

The Employment Tribunal’s judgment states that there are different protected characteristics under the Equality Act. However, it’s not specifically stated that one protected characteristic takes precedence over another. In Forstater, the Employment Appeal Tribunal emphasised that it is important to bear in mind that the Equality Act does not create a hierarchy of protected characteristics. That wording, or anything like it, is simply not in the Forstater judgment.

Additionally, the judgment referred to in paragraph 792 erroneously states that the Supreme Court endorsed the EAT judgment as "comprehensive and impressive". However, that phrase is in reference to the Forstater case, not related to creating a hierarchy of protected characteristics, thus making this paragraph incorrect.

In paragraph 793, there is another non-existent quote from the Supreme Court in Lee and Ashers. When checking this quote, you might find that AI could generate fictitious quotes if you simply ask if this judgment stated something.

The text from paragraph 793 reads, "that one protected characteristic does not take precedence over another is also consistent with how different human rights are addressed." In the judgment, it states, "the rights to freedom of thought, conscience, and religion, and to freedom of expression,” but the quote should refer specifically to those in the judgment. The judgment shouldn’t have a quote that it simply cant substantiate.

It appears that either sections have been fabricated or enhanced through AI, or the tribunal is twisting what was in that judgment. Either way, they cannot quote the Supreme Court in this respect, as it makes it look like the Supreme Court actually said this when, indeed, it did not.

In paragraphs 803 and 804, there is also a selective quote from For Women Scotland, omitting significant context that changes the interpretation. Anna points out that the tribunal can correct quotes under the slip rule, but you cannot amend entire paragraphs post-judgment.

As we proceed through these paragraphs, it's evident there has been selective quoting to create differing meanings and to hastily conclude judicial interpretations that don’t exist. This is simply unacceptable and is the kind of error one wouldn’t expect in GCSE law, in short.

Additionally, in comparing paragraph 808 with its actual quotes, it’s clear selective quoting is once again being used to create a misleading narrative. The Supreme Court's commentary, which details practical difficulties, was deliberately left out, which significantly alters the entire meaning.

I would say that the entire judgment warrants appeal, as it is fraught with errors. While Sandy Peggy did succeed on some of her claims, others may well be appealable. It is indeed an extraordinary situation. Please let me know what you think in the comments below, and don’t forget to watch the 'Black Belt Secrets' video alongside this as well.