top of page

Artificial Intelligence - LinkedIn Lightbulb Series - March 2024



Welcome to my Legal Lightbulb series for March 2024, where I’ll be tackling the legal questions around AI. What would you like to know? 🤖


Artificial intelligence has come on leaps and bounds in the last couple of years and it seems we spent most of 2023 either excited or terrified by it. I’ve heard people gleefully spew about how transformative AI has been for them - saving them time, helping them become more organised and taking tiresome tasks from their to-do lists. 😃I’ve heard others worry that AI might make their whole career redundant, that it’s lowering the quality of content and conversation and that it could be very dangerous for society.


😱There’s no doubt that there are big questions to be asked about the ethics of AI and the role it’s going to play in our future. However, there are also important legal considerations. Especially for business owners.Right now the legal system is very busy working out how to legislate artificial intelligence, particularly where there is conflict with other laws such as intellectual property, data protection, etc.


⚖️So this month I’m going to highlight some of the most common and interesting questions around AI and the law. 🔍This is going to be highly important for anyone using AI, or considering using AI in their work (even if it’s only the tiniest bit). So, if you’re not already connected with me, please do so now.


Q. If I put content into an AI processor (like ChatGPT) can it reuse my content? 🤔


A. Usually, yes. It’s a standard part of the T&Cs for most AI platforms. So if you’re inputting into an AI word processor to rewrite, comment on, or teach the AI about the style or approach you need, etc, then you’ve effectively permitted it to use this and to redistribute it. It may use this to assist other users, it may regenerate the content or the information as an example to support an argument or theory. 🔄


For instance, it would be entirely possible for a fictional crime story to be cited by AI to support a legal case, because the software may not be able to identify it as fiction or understand that that would matter. 🕵️‍♂️


Think very carefully before inputting anything into AI and, if it isn’t your original content, you should first acquire permission from the client/copyright holder. Anything that needs to be kept proprietary probably shouldn’t be input. 🚫



Q. Could AI Bias Impact My Advertising?


A. Because AI indiscriminately mines data, sometimes it can present responses that carry bias. In some cases, it has even been downright discriminatory. Algorithmic bias in AI threatens a few aspects of law and industry (and we’ll return to these through the series), but breaching advertising guidelines as a result of AI bias is one worth serious reflection.


🤔AI language processing tools, such as ChatGTP, learn from data. Unfortunately, some of this may be unbalanced or discriminatory. This might lead an AI to produce work carrying biases based on gender, race, age, health, religion, disability, sexual orientation or other characteristics. 🙅‍♂️🙅‍♀️This could be a big problem if a business is using AI to write marketing content, design adverts or develop promotional campaigns.


📢When using AI to generate content for public use, without a mind for filtering out prejudice and unconscious bias, we risk putting out materials that overlook minority groups/marginalised people or feed stereotypes.Whilst bias in advertising may not lead to legal action you could find yourself in trouble with the ASA (Advertising Standards Authority). It could impact your brand reputation as well as put people off buying your products or services.I recently heard a YouTube advert for some kind of home security kit where they claimed whether ‘you are an engineer or just a housewife, you’ll be able to set this up’.


Now, I would bet big money that this advert was AI generated - at least the script was - because I can’t imagine anyone would actually write that, but - someone must have signed it off. So, if you’re using AI in your promotional marketing then make sure you're scrutinising the output, not only for quality, but for bias.#AI #AIBia #Advertising #SmallBusinessPromotion #Stereotypes #ASA #AdvertisingStandards #Ads #AIGeneratedMarketing


Q. Can I Legally Use AI To Do My Work?  🤖


A. This is not quite straightforward. If you are employed and your employer has permitted you, or even asked you, to use AI to assist in your work, then you shouldn’t have a problem in terms of your employment. However, the extent to which you use AI and how you use it could have an impact on legal compliance and ownership.🤔


In the UK, copyright still belongs to the person or company creating the work even if they are using AI assistance, but this is not true internationally. 🌐


In the US, work cannot be copyrighted if it isn’t considered to be the work of a human creator and this is also similar in regards to some cases in European law. Something worth considering if your company operates internationally is that ownership over the work produced may be contested if using AI. Meaning you may not qualify for ownership of the work under some jurisdictions if it has been computer-generated.


It is important to understand the limitations an employer has on the use of AI to assist employees in their work. It is best to get permission for any use of AI in your work, should it affect your employment or conflict with company policies around non-disclosure, security, etc.


There is also the question of whether work produced by AI is of the quality expected in your deliverables. Even if AI use is permitted, computer-generated work may not be considered of the standard expected and this could then compromise your employment.


Employees must ensure their use of AI complements the quality of work and does not compromise it. To achieve this, and to be sure your employer is supportive of your approach, honesty around the use of AI-assisted technology in the workplace is paramount.


Employers should consider creating policies around AI usage so that employees have clarity about how they may or may not use it in their daily work. 📝



Q. If I Use AI In My Business Do I Own The Copyright? 💼 


A. In the UK, yes. 🇬🇧 


In the US and EU - maybe not. Probably not. 🙅‍♂️


The CDPA (Copyright, Designs and Patents Act) will accept the author as the human behind the work, even if it has been generated with or by AI. For example, if a company has used ChatGTP to create all their marketing materials then they are still legally owned by the company. Unapproved use of these materials by any other person or entity would be copyright infringement. However, this is not the case in many other countries whose laws and regulations do not assign copyright to computer-generated works. 💻 



Q. Can I Share Sensitive Information With AI ChatBots? 🤔🤖


A. Even the creators and businesses behind popular AI tools, including ChatGPT and Bard, caution users against inputting private or confidential information into AI tools and chatbots. 🚫


🔐AI learns from the content put into it, including yours, and there's always a risk this could appear elsewhere online, even if in a different form. Manual reviews of AI prompts also take place so the customer experience/service can be checked and reviewed by human staff. 👩‍💻👨‍💻


Furthermore, AI prompts and portions of chat histories have already been leaked online leading to crucially confidential information entering the public domain. 🌐

Though most AI tools endeavour to keep your prompts confidential, they cannot guarantee it - at least not yet. Therefore, personal data (especially on clients or dependents) should never be shared with AI. Neither should sensitive information such as company finances, confidential code, forecasts and plans, etc. 📉💻


Businesses allowing the use of AI should consider putting policies in place for staff and providing proper training to avoid data breaches occurring through AI use.



Q. Are there implications for using AI in recruitment?


A. Businesses should be wary about using AI in recruitment processes. Algorithms in AI may make decisions that are biased in nature. For instance, systems using AI to screen candidates may disregard nuanced skills or experience. ⚠️


AI may also use personal data to automate decisions that can breach GDPR and raise privacy issues. Even if actioned by a human person, AI systems may be discriminatory. 🛑


Of course, AI may reduce time spent screening CVs and, in the future, AI may even help to reduce or eliminate discrimination in recruitment. Yet, in these early days, HR and recruiters should ensure that when AI is used in recruitment processes, decisions made are vetted by a person/people who understand GDPR and equality rights and protections.



Q. Can using AI put me in breach of copyright? ⚠️


A. In the UK anything created by you will be automatically under your copyright, even if AI has assisted in the work. However, if the published work contains copyrighted material from any other person or organisation then you could find yourself in breach of copyright.


In generating content, AI scans a huge pool of data and uses this information to create your content. Sometimes, it may not repurpose or adapt this content as much as is required to ensure it’s not unlawfully trespassing on intellectual property. 🛡️


There have already been a number of legal cases raised where an author has accused AI of lifting their content for republication or accessing and utilising content without paying for the privilege. Such practices could severely impact creatives and academics publishing written content. This is mostly an ongoing dispute between content creators and AI tool creators but could become an issue for those using content generated by AI that may be considered a copyright infringement. This would apply to AI image creation also.


It’s important to recognise that there is a fair use clause in copyright law that allows for pieces of copyrighted material to be used for certain purposes, including criticism, commentary, news reporting, teaching, scholarship, or research. 📚


If you’re using AI to generate content or conduct research, one thing you can do to protect yourself is to ask ChatGTP to cite its sources. Knowing where information was gathered from will be important in shielding yourself from breach of copyright when using AI tools.



Q. Can AI be used to create legal contracts? 📝


A. The simple answer is yes. The question we should be asking though is ‘Should’ AI be used to create legal contracts?


I asked ChatGTP if it could write legal contracts for a business and this was the response:


“I can certainly help you draft a legal contract, but please note that I'm not a lawyer and cannot provide legal advice. It's always recommended to have a qualified attorney review any legal documents to ensure they meet your specific needs and comply with relevant laws in your jurisdiction. If you have specific requirements or details you'd like included in the contract, please provide them, and I can help generate a draft for you. However, keep in mind that I can't guarantee the accuracy or legality of the document without professional legal review.” 🤖


Accuracy in legal documents and contracts is essential, otherwise your contract may not protect your business. For instance, if a dispute is raised and your contract is found to be unreasonable or unclear then it will be considered void. 🛑


Although it may be tempting to save money and use AI to help with legal contracts, it might cost you far more if mistakes are made. Unless you have a legal background and are able to properly review contracts, using AI in drafting legal documents could put your business at risk.


Working with a commercial solicitor to draft contracts may be more affordable than you think. Get in touch for a free consultation and we can discuss your requirements. 📧



Q. Do I Need To Tell Clients If I’m Using AI In My Business? 🤔


A. There’s no reason to share the utilisation of AI in your business with clients unless you’re using it to:


  • Collect, store, manage and/or process their data 📊

  • Create work for them 🖊️


Although there is no legal requirement in the UK currently to disclose the use of AI, when it comes to data protection it’s important to be transparent. Therefore, if AI is aiding you through CRM systems or other processes dealing with personal data, you may need to disclose this in your privacy policy to ensure you’re compliant with GDPR guidance. 💼


If you’re using AI to aid you in creating work for a client, again there is no legal framework for navigating this (yet). However, it may be argued that you're using a third party and potentially sharing the client’s personal information with the processor in the course of your duties for the client. In many circumstances, such actions would need to be pre-agreed with the client. 


Therefore, if you’re using artificial intelligence to produce work as outlined in a contract (typically in terms of business, services agreements and/or terms and conditions) then it would be wise to address this in the contract. 📝 


If AI use is declared and pre-agreed then you should avoid related disputes, so long as you are clear in your legal agreement(s) about when and how AI might be utilised.


Concerned about the legal and IP risks of AI usage in your work for clients? DM me with any concerns or queries you may have and I’ll endeavour to advise you.



Q. What Should I Consider When Using AI In Product/Service Descriptions? 🛍️ 🛒


A. AI is very impressive - no doubt. Language processing tools like ChatGPT can swiftly transform a little information into comprehensive text and, when done well, it can be undetectable as AI scripted. However, these tools have a tendency towards artistic licence and they’re using examples from a range of inputted data to enhance your key information. This may mean that product descriptions written by AI may not accurately reflect the properties and scope of your products/services. 💼🔍


So what’s the harm in that?


Well, it may threaten consumer rights and put your business in jeopardy. Under the UK Consumer Rights Act, products and services must reach the quality standards described by the business. If you fail to meet these standards, as laid out by the description of your products or services, the consumer has the right to a refund. Using AI to assist in writing these product descriptions, therefore, can be risky.


If you wish to use AI to aid your product or service descriptions then you must be meticulous in reviewing the content against the service/product to ensure it truly reflects capabilities, quality and reasonable customer expectations.


For more about consumer rights, please refer to my article - What Businesses Need To Know About Consumer Rights









Comments


bottom of page