This article provides an overview of some of the potential legal considerations and risks for employers to contemplate when seeking to deploy artificial intelligence in the workplace. It mainly focuses on employment law considerations, but also touches on other key issues such as data protection and generative artificial intelligence.

Most organisations already use artificial intelligence (AI) in some form or another, including to support workplace and HR functions – whether it’s using email spam filters, having auto-complete functions, or using preliminary CV scanning at the recruitment stage. Recent technological advancements, and the ease of access to generative AI (GenAI), is accelerating the pace of adoption of AI and new AI-based products, and there are a whole host of attractive AI solutions available to improve workplace efficiencies.

This note does not intend to cover all aspects of AI and the associated legal risks, but it aims to give you a brief overview of some of the potential legal considerations and risks for employers seeking to deploy AI in the workplace. The requirements of each organisation will vary, and organisations may also be subject to sector regulatory requirements, so in this note we have reflected on some of the core considerations, which will be common to all organisations.

What laws regulate AI at work, and will new laws be introduced?

There is currently no specific UK legislation regulating the use of AI generally, nor the use of AI in the workplace. The House of Commons Library recently published a report which reviews the interaction between AI and the current employment legislation, which can be accessed here. Existing laws relating to data protection, equality and common law principles such as the mutual duty of trust and confidence, will all be relevant when considering how AI can be used at work, but there is no single piece of legislation governing its use.

In light of the UK Government’s White Paper, released in March 2023, it is unlikely that any new UK legislation will be introduced on this topic in the immediate future. Instead, the government will support existing sector regulators to control these technologies in their sector using a principles-based approach to AI regulation, focussing on the following guiding principles (which will not be statutory): safety, security and robustness; appropriate transparency and explainability; fairness; accountability and governance; and contestability and redress.

As such, it currently falls to existing legislation to govern the use of AI in the workplace. Given AI has the potential to touch on almost every aspect of the employment relationship, it’s important that organisations are familiar with the current legal landscape and how this interacts with AI at work, to ensure they can maximise opportunities and mitigate risk. Organisations with a workforce based outside of the UK will also need to consider the legal landscape in each of the different countries of operation.

What legal risks do we need to consider if we wish to introduce AI into the workplace, and how do we manage such risks?

There is no doubt that AI can be extremely positive and help businesses operate more efficiently, but there are a number of relevant risks to consider. We’ve highlighted a few of the key ones below.

Discrimination

Perhaps the most commonly reported criticism of AI, in an employment context, is the discrimination risk. It’s widely acknowledged that AI tools can exhibit biases – this could be because the training data an AI system learns from might be unbalanced (i.e. if the majority of the data has come from one predominant group this could adversely impact groups who are not as well represented) or even reflect past discrimination. Discrimination may also arise from the way in which AI is used or because an employee hasn’t reviewed the output sufficiently.

There’s the potential for AI to be discriminatory on the basis of any protected characteristic. For example, it’s been widely reported that in 2014, Amazon developed a recruitment tool, which aimed to screen CVs to identify the strongest candidates, but it had to be shut down because it discriminated against female candidates.[1] Indeed, Uber Eats’ use of facial recognition technology to verify the identity of workers when they log onto a shift has been the subject of allegations of race discrimination, after concerns were raised that the system didn’t work fairly for people of colour and failed to recognise them.  Furthermore, a recent study by University of East Anglia, suggested that the ChatGPT model (built by OpenAI using large language models), had a ‘significant and systemic’ tendency to return politically left-wing responses when questioned.[2]

In terms of liability, employers can be vicariously liable for the actions of their employees in certain circumstances – this would extend to the use of AI. Further, if an employer adopts AI systems which are inherently biased or discriminatory, they will be liable for any resulting discrimination.

It’s important to evaluate the potential discriminatory effects of any AI tool you are considering using before introducing it. The use of AI may well amount to a ‘provision, criterion or practice’, which, if it has a disproportionate impact on someone with a protected characteristic, could result in indirect discrimination claims. Further, if there’s a blanket use of AI in disability cases, there may well be claims for failure to make reasonable adjustments and/or discrimination arising from disability.

It’s therefore crucial that an organisation understands how the AI model operates and the intended outcomes before it is introduced; it is also important to ensure that the data sets which will be used to train the AI are verified and biased data has been removed.

Contracts with AI providers should specify the requirements for the training data and for testing the AI to address bias before it is deployed.  They should also set out the steps the supplier will take if, after deployment, it becomes apparent that the AI displays bias or provides an output which causes discrimination.

Appropriate training on how the AI system works and how to interpret the output should be provided to any individual who is going to use it. We’d also recommend that you introduce a policy which governs acceptable use of AI: setting out clear guidelines and expectations should help mitigate risk. For more information on introducing a Generative AI policy, see our recent article. Finally, it goes without saying that any decision produced by AI must be subject to careful human review by an individual with appropriate skills and qualifications.

Lack of transparency

Transparency is a key AI guiding principle as well as one of the seven GDPR data protection principles (see below). Reliance on AI can result in a lack of transparency. The transparency ‘void’ makes it harder for employers to justify decisions and harder for employees to challenge those decisions. Indeed, in some cases, employees may not even be aware that AI has been used in their employer’s decision-making process.

If an employer is using AI in the workplace, they need to be clear and transparent about what AI they are using, when they are using it, for what purpose they are using it and the impact of its use on their employees. There must be a legitimate reason for the use of AI, and it must be proportionate. In addition to the data protection risk, failing to tell employees about the use of AI in a way which impacts them undermines the mutual duty of trust and confidence between an employer and employee and is likely to lead to significant employee engagement issues. Indeed, in serious cases, it could lead to a fundamental breach of contract resulting in an employee resigning and claiming constructive dismissal.

Once AI has been introduced, an employer will also need to be able to justify any decision which has been made been made in reliance on AI. This will inevitably involve the employer (i.e. the individuals who will be making decisions on behalf of the business) having a sufficient understanding of the AI model to explain the decision. If a manager is not able to understand how the AI operates, they are not going to be able to explain their decisions, nor demonstrate that the decision was rational, fair and made in good faith. Again, this feeds into the need for mutual trust and confidence in the employment relationship.

To reduce the risk of unfair decisions, an individual with appropriate skill and expertise should always have the final oversight and say on any decision. It goes without saying that the individual should also have training on how the AI works, and how to interpret the outcome data. Being unable to explain or justify a decision is not only going to damage the relationship with an employee, it may also lead an employee to believe that a decision has been for other reasons (which may be discriminatory or otherwise legally unfair) in the absence of a rational explanation.

In practice you will need to ensure that there is a user manual for managers, which is kept updated as the model develops. It will also be important to ensure that a record-keeping repository is maintained to help explain decisions if they are subsequently challenged.

Changes to the workforce

The introduction of AI has the potential to fundamentally change the way the workforce operates.

In some cases, AI might lead to roles being eliminated and a redundancy process ensuing. On the other hand, it could lead to a recruitment drive with new jobs being created or staff being upskilled. It may also lead to roles being amended over time.

It’s therefore not out of the question that restructures and reorganisations may be needed as AI technologies become embedded in organisations. This will need to be handled fairly and sensitively, and meaningful consultation will need to be undertaken to reduce legal risk.  Again, engaging with staff (and unions if relevant) and being transparent about the introduction of AI should help tease out any issues and help balance interests.

Impact on staff wellbeing

There are legitimate concerns about the potential for AI to have a detrimental impact on employee health. There is no doubt that the introduction of AI can be positive and aid employees in their roles, particularly by making processes more efficient and reducing the administrative burden. However, there are concerns that if every administrative aspect of a role is removed, work could intensify in a way which may risk mental or physical health. Further, if employees are constantly monitored, tracked and micro-managed by technology, this can lead to increased stress and have a negative impact on mental health. This could become even more pronounced if technology trackers are used on portable devices, which could severely impact work-life balance. There is also the widely publicised concern that the widescale introduction of AI could start to render some human held jobs obsolete, which of course, leads and has already led to some fearing for their livelihood.

Given the focus on mental health at work and employee wellbeing, this will need to be handled sensitively by employers. Again, engaging with staff (and unions if relevant) and being transparent about the introduction of AI should help tease out any issues and help balance interests.

Does AI present any particular data protection issues?

Most AI used in relation to the workforce involves the processing of a significant volume of personal data, including special categories of personal data.[3]  Organisations seeking to deploy AI should therefore pay careful attention to data protection requirements well in advance of deploying AI technology. The seven GDPR principles (lawfulness, fairness and transparency; purpose limitation; data minimisation; accuracy; storage limitation; integrity and confidentiality (security); and accountability) apply whenever personal data is processed and these principles overlap with many of the AI guiding principles.

The processing of personal data by AI is generally considered to present a higher data protection risk, and the use of AI systems which process special category data and those involving foundation models,[4] can present even greater data protection risks so careful analysis of the risks should be carried out before deployment. The ICO recommends that a data protection impact assessment (DPIA) should be carried out whenever an AI system involving the processing of personal data is deployed. A DPIA will enable you to assess the risks and how those risks can be mitigated. The Information Commissioner has produced a helpful guidance and a toolkit[5] to help organisations seeking to deploy AI.

Each AI system will present different levels of data protection risk, DPIAs should be carried out for each one and revisited on a regular basis throughout the lifetime of the AI system as the risks to data protection may evolve over time. You should consider how the operation of the AI system can be suspended if it transpires that there is a risk of or actual breach of data protection.

Legitimate Interest Assessments are also of critical importance to ensure that there is a lawful basis for the processing of the employees’ personal data and, as is widely acknowledged, there can be risks in relying on consent given the power imbalance in the employer-employee relationship.

Consistent with the AI transparency requirements, information needs to be given to your employees about the processing of their data so you will need to consider what updates may be required to employee policies and manuals as well as whether any further notices are required at the point of the collection of data.

Additional rules apply where an employer carries out solely ‘automated decision-making’ or ‘profiling’ that has legal or similarly significant effect on them.[6]  Under GDPR, solely automated decision making is generally prohibited, save in limited circumstances.[7] This topic is outside the scope of this note and we recommend that advice should be taken in any circumstances where there may be automated decision-making.

AI systems which involve biometric data, such as facial recognition systems present further risks and require additional consideration. The ICO has issued draft guidance on the use of biometric data.[8] For further information about the risks involved please also see our article on use of biometric technology.

In view of the complexities and inherent data protection risks in the context of using AI in the workplace we recommend that your Data Protection Officer should be involved from an early stage.

Does Generative AI present other risks?

GenAI, such as ChatGPT and Google’s Bard, which can be used to generate content and provide ‘human like’ answers to complex situations, presents many opportunities and potential efficiencies, which organisations will be keen to explore. However, due to the manner in which GenAI operates by inferences or predictions and the immense quantities of data involved in the building the models which underlie GenAI, the legal risks are even further amplified.  Very careful consideration of the risks, mitigations and safeguards should be conducted before GenAI is deployed in connection with HR functions.

We will comment further on the subject of foundation models, large language models and GenAI in future but in the meantime, for more information about introducing a Generative AI policy, you can read our article.

For advice on the data protection implications of AI please contact our Data Protection & Privacy team and for advice on the particular issues to take into consideration when contracting for the procurement of AI please contact our Technology & Innovation team.

If you’d like to discuss any of the employment/HR related issues raised in this note, please contact Lynsey Blyth.

This article is for general information only and does not, and is not intended to, amount to legal advice and should not be relied upon as such. If you have any questions relating to your particular circumstances, you should seek independent legal advice

[1] The AI algorithms were trained to observe patterns in the CVs submitted to the company over a 10-year period. As most of the CVs came from men, the algorithm spotted male preference and then discriminated against female candidates, penalising CVs which included the word ‘women’s (such as ‘women’s chess club captain’).

[2] The findings were published on 17 August 2023 in the journal Public Choice and were based on research carried out in Spring 2023, using version 3.5 of ChatGPT – more details of the study can be found here.

[3] Personal data revealing racial or ethnic origin; political opinions; religious or philosophical beliefs; trade union membership; genetic data; biometric data (where used for identification of a person); data concerning health; data concerning a person’s sex life; and data concerning a person’s sexual orientation.

[4] Foundation models and large language models are AI’s trained on huge quantities of data and are often used as the base for building generative AI.

[5] https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/artificial-intelligence/guidance-on-ai-and-data-protection/

[6] ‘Automated decision making’ is the making of a decision, about an individual, based solely on automated means without any human involvement and ‘profiling’ is any form of automated processing of personal data to evaluate certain things about an individual.

[7] https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/individual-rights/individual-rights/rights-related-to-automated-decision-making-including-profiling/

[8] https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/guidance-on-biometric-data/data-protection-requirements-when-using-biometric-data/#doweneed2

The Firm is celebrating its latest Legal 500 results, which include 29 team and 39 individual rankings for 2024. This year’s results see five new individual rankings, two improved team rankings, and retention of 9 Tier 1 rankings in strategic areas for Michelmores.

We have been ranked as a Tier 1 firm for Agriculture & Estates (South West) and Tier 2 for Agriculture & Estates (London), with Partner Rachel O’Connor from the Firm’s Agriculture team having climbed up the rankings as a Next Generation Partner.

Ian Holyoak, a Partner in the Firm’s Commercial team, has been recognised as a Leading Individual in the Energy, Project Finance & PFI (South West) rankings joining our four other top ranked lawyers in this area.

Philippa Collison, an Associate in the Firm’s Commercial team, has also been listed as a Rising Star (South West) for her expertise in brand protection and IP transactions. Chloe Vernon-Shore, a Partner in the Firm’s Commercial team, was newly listed as a Next Generation Partner.

New additions to the rankings for Michelmores also include Rachael Lloyd, who was promoted to Partner in May 2023 and has been ranked as a Next Generation Partner in the Employment law sector.

The Firm also increased its rankings in the IT and Telecoms London region climbing to Tier 4 showing recognition of Michelmores services to clients across the technology sector, covering a full spectrum of commercial, corporate, contentious and regulatory matters in the City market.

For the full Michelmores results, please visit The Legal 500 website.

The purpose of the Legal 500 UK Solicitors 2024 rankings is to help in-house lawyers and legal teams find the right advisors. These rankings reflect the comprehensive analysis undertaken by The Legal 500’s legal directory research team. Along with robust assessments of law firm submissions, they also conduct interviews with thousands of clients who play a key role in informing the results.

For more information about Michelmores or to get in touch, please visit our website.

High Speed Two (HS2) was the Government’s proposal for a new, high speed north-south railway. Phase Two of HS2 was intended to connect the West Midlands to Manchester, Leeds and beyond.

Early preparations for Phase 2 have seen the compulsory acquisition by HS2 Ltd, the public body responsible for developing HS2, of thousands of homes, properties and acres of land, from which owners have been ejected.

The Prime Minister’s announcement that Phase 2 of HS2 from Birmingham to Manchester will be cancelled has caused consternation in many quarters, not least among those land and property owners, whose property will no longer be required for HS2.

There is a glimmer of hope for those land and property owners, who may be offered their land back in accordance with a set of policies set out for this purpose.

Crichel Down Rules

These are non-statutory rules dating from the 1950’s that relate to the offer back to previous owners of land acquired from them by, or under threat of, compulsory purchase. They broadly set out the policy which will usually apply to this situation.

HS2 Land Disposal Policy

HS2 recognised that not all property compulsorily acquired for HS2 would be used permanently for the construction and operation of the new railway. For example, land acquired to be used for access and construction, would be surplus to requirements once the railway is complete.

HS2 therefore developed a Land Disposal Policy, compiled upon the basis of the Crichel Down Rules, containing Guiding Principles, governing how land that was compulsorily acquired, but is no longer required, should be dealt with.

The cancellation of Phase 2 of HS2 means it is likely that more former owners and occupiers of land acquired for the scheme will stand to benefit from the HS2 Land Disposal Policy than if Phase 2 had proceeded.

HS2 Land Disposal Policy terms

The general rule is that where government wishes to dispose of land to which the HS2 Land Disposal Policy applies, former owners will be given first opportunity to repurchase the land at current market value.

Some of the points to be aware of include:

Qualifying Interests: To qualify for the offer back of an interest you must be a holder of one of the following Qualifying Interests:

• former freeholders of the whole or part of a site; or
• (where the former freeholder does not wish to buy back the site) the former leaseholder who had a lease of the site with an unexpired term of at least 21 years;
• the successors of anyone who would have fallen into either of the above categories had the property not been acquired; or
• where there was fragmented ownership of a site at the date of acquisition, a consortium of former owners who have indicated a wish to purchase the land collectively[1]; and
• in certain circumstances, a sitting tenant of a dwelling[2].

The Secretary of State may however decide that the land should not be offered back where, among other reasons:

• any works have materially changed the character of the land since its acquisition, such as where a building with land was acquired and then demolished, but only part of the land has become surplus to requirements;
• agricultural land that has been severed is no longer capable of being farmed economically;
• in the view of the Secretary of State it makes sense to pool the land with adjoining ownerships in a joint disposal;
• the site is required for environmental mitigation, and where the former landowner is unwilling or unable to accommodate those requirements for recreation of community facilities or wildlife habitats;
• former owners are not prepared to pay the market value of the site or are not prepared to offer terms that the Secretary of State considers to represent best value having regard to all the circumstances; and
• the Secretary of State considers that in the public interest the land should be transferred to another Public Body with compulsory purchase powers.

The Crichel Down Rules contain prescribed procedures and timescales for communication by the disposing public body, in this case HS2, of an intention to offer land back to the former owner. It is important that all deadlines contained in the communications are met.

HS2 Limited and the Secretary of State can be expected to comply with their obligations under the Crichel Down Rules and the HS2 Land Disposal Policy and all associated Guidance where it is decided that the land acquired is no longer required for the purpose for which it was compulsorily acquired. The precise purpose of acquisition, and the extent to which it is deemed the land is no longer required is perhaps likely to be contested in some circumstances.

Our lawyers have experience in advising land owners and public bodies in the exercise of their rights and obligations under the Crichel Down Rules and infrastructure project land disposal polices, as well as in related contested court or tribunal proceedings

[1] N.B. the consortium must be established before the wish to purchase is indicated, and in sufficient time to enable deadlines for notification to be met.

[2] N.B. this does not apply to agricultural units.

We are pleased to announce that Daniel Eames, an International Family Law expert and head of the Family team at Michelmores, has been featured in The International Comparative Legal Guide – Family Law, 2024 edition, as author of the England and Wales Chapter.

The 2024 edition covers key family law issues across worldwide jurisdictions, including divorce, financial remedies, marital agreements, cohabitation, child maintenance, child arrangements and international matters.

ICLG is one of the highest-ranking legal reference websites, and a leading comparative authority on family law across many countries worldwide.

Daniel is highly experienced in all aspects of Family Law and is recognised nationally as an expert in International Family Law including by independent legal directories – Chambers & Partners as a top tier practitioner in their High Net Worth Guide and by the Legal 500 UK.

Read Daniel’s ICLG chapter here and get in touch via our website for more information.

Merry and Dyer (Joint Liquidators of Safe Depot Ltd (in Liquidation) and another v Esa [2023] EWHC 2011 (Ch), [2023] All ER (D) 19 (Aug)

Introduction

Section 214 of the Insolvency Act 1986 (the Act) provides that once a director of a company concludes (or should have concluded) that there is no reasonable prospect of the company avoiding an insolvent liquidation or insolvent administration, they have a duty to take every step that a reasonably diligent person would take to minimise potential loss to the company’s creditors. Section 214 also provides that if, after the company has gone into insolvent administration or liquidation, it appears to the court that a director has failed to comply with this duty, the court can order that director to pay a contribution to the company’s assets.

The lack of company books and records however can mean that it is difficult for office holders to show:

1. what the company’s financial condition was on any given date and at what point insolvency was inevitable; and
2. when a director did, or should, have realised the point at which a company became insolvent.

In this case the Liquidators were able to provide sufficient evidence to enable the court to make an order under section 214, in the context of a director conducting an ‘informal winding up’ by transferring out the profitable elements of the business, whilst the company continued to incur liabilities.

Background

Safe Depot Ltd (the Company) operated as a provider of storage space across three sites in the northwest of England. At the relevant times prior to its winding up, the respondent director was the sole registered director and shareholder of the Company.

On 24 July 2017 the Company was wound up on a petition presented by the landlord of one of its premises. The Joint Liquidators appointed over the Company (the Liquidators) identified that:

• The Company suffered from cash flow problems from 2014, and eventually serious financial difficulties, such that it was cash flow insolvent by April 2016 and balance sheet insolvent by September 2016. In September 2016 the Company had instructed debt advisors with a view to putting the Company into creditors voluntary liquidation, however this never materialised.
• Despite its insolvency, the director caused the Company to dispose of various assets of each of its storage sites. This included the transfer of a customer list to a competitor, and also the transfer of leases and book debts to a connected company of which the director was also sole director and shareholder (together, Transfers).
• During the period between September 2016 and the winding up order being made in July 2017 the Company’s creditors had increased considerably (according to the Liquidators, by at least £433k).

The Liquidators therefore issued proceedings against the director, on the grounds that the Transactions constituted transactions at an undervalue, or alternatively preferences, or in the further alternative, transactions defrauding creditors. The Liquidators also alleged that in causing or permitting the Transactions the director acted in breach of his duties to the Company.

The Liquidators further claimed that the director had engaged in wrongful trading for the purposes of section 214 of the Act because from September 2016 the director knew or ought to have known that there was no reasonable prospect of the Company avoiding insolvent liquidation, and yet allowed it to continue trading.

The court’s decision

The judge held that the Company was cashflow insolvent from April 2016 and in any event by August 2016, at which point it was probable that the Company would go into insolvent liquidation. This enabled the judge to make the following findings.

Antecedent transactions, breach of duty and determination of value

The judge determined that there was no evidence of any consideration having been given for the Transfers. It therefore followed that the Transfers amounted to a transaction at undervalue effected by the director in breach of his duties to creditors.

The judge considered however that the Liquidators had not been able to present evidence as to the value of certain elements of the transfers, including the customer lists, goodwill and leasehold property, and therefore those elements of the claim were dismissed. There was sufficient evidence of the value of the Company’s debt book to enable the judge to make a determination on the value of that aspect of the claim.

The judge also did not find that the evidence was quite enough to establish that the transactions were entered into for the purpose of putting assets beyond the reach off creditors amounting to a transaction defrauding creditors under section 423 of the Act.

The judge went on to find that there was no basis for granting the director relief from the consequences for his breach of his duties under section 1157 of the Companies Act 2006. The director should not be allowed to enjoy the benefits of the Transfer (in his capacity as sole director and shareholder of the recipient company) at the expense of the Company’s creditors.

Wrongful trading

The judge found that despite the director’s contentions that the Company ceased to trade in July 2016, it was demonstrable that it continued to incur liabilities. The director knew, or ought to have known, by the end of September 2016 that there was no reasonable prospect of the Company avoiding insolvent liquidation. The judge noted that the director’s conduct amounted to an ‘informal winding up’ where the profitable parts of the Company’s business were passed to the connected company.

The judge therefore held that the requirements of section 214 of the Act were made out as at 29 September 2016 and by that date the Company should have stopped incurring liabilities and been placed in an insolvency process. There was no evidence to suggest that the director took any steps at all to minimise losses to creditors. It was therefore ordered that the director would be required to make good the losses caused by the Company continuing to trade after 2016.

Conclusion

This case demonstrates that, despite the difficulties in proving wrongful trading and the rarity in which orders against directors for wrongful trading are made, the court will be prepared to make orders under section 214 of the Act when office holders are able to produce documents that show the required information in sufficient detail.

It is notable that the relevant timeframes did not include the period during which the temporary provisions of the Corporate Governance and Insolvency Act 2020 (CIGA), suspending liability for wrongful trading, applied (1 March 2020 to 30 June 2021) and that reported cases where the Court has been required to determine the personal liability of directors during time periods that include that time have so far been few and far between.

The decision in HMRC v A Taxpayer [2023] UKUT 182 (TCC) was handed down on 28 July following an appeal by HMRC in response to the First-tier Tribunal (FTT) decision in May 2022. This is the first case of its kind to examine the meaning of “exceptional circumstances” for the purposes of the Statutory Residence Test (SRT) under paragraph 22(4) Schedule 45 Finance Act 2013. Whilst the FTT decision provided some hope for a broader meaning of exceptional circumstances, taxpayers will be concerned to note the much narrower construction adopted by the Upper Tribunal.

The facts of the case

The taxpayer left the UK to live in Dublin on 4 April 2015, following UK tax residency in at least one of the previous three tax years. She had received legal advice regarding the SRT and her move to Dublin was motivated by this. She was aware that to avoid being UK tax resident for the 2015/16 tax year, she could spend no more than 45 days in the UK. However, she visited the UK for a total of 50 days during the course of the tax year.

She filed her 2015/16 self-assessment tax return on the basis that she was non-resident and that five of the 50 days she was present in the UK was due to exceptional circumstances. The relevant stays in the UK were split over December 2015 and February 2016. HMRC subsequently opened an investigation into her return and rejected her claim of exceptional circumstances for the five days, issuing her with a tax bill of over £3 million. The taxpayer appealed the decision, and the case went to the FTT.

During the course of the FTT hearing, the taxpayer put forward the case that she had been in the UK due to her twin sister’s alcoholism, depression and threats of suicide. She also claimed that her sister’s two minor children had been neglected and the house was in an unliveable state. The taxpayer’s defence was made up of two prongs:

1. Her sister had threatened suicide which constituted “exceptional circumstances beyond her control” and the taxpayer was unable to leave the UK until her sister was “in a place of safety”
2. Her sister was unable to care for her children and the taxpayer was unable to leave the UK until care arrangements had been made for the children

Whilst the FTT rejected the first ground, it accepted the second on the basis that the combination of care requirements for the taxpayer’s sister and her children constituted exceptional circumstances with a heavy weight being placed on the taxpayer’s moral obligations towards the children.

The judgment

HMRC’s appeal was successful, and the decision made by the FTT was overturned by the Upper Tribunal (UT), with the taxpayer being found UK tax resident for 2015/16. The UT noted that the taxpayer was aware of her requirement to keep a clear record of her days in the UK following the legal advice she had taken and, also, that she would be considered UK tax resident if she stayed in the UK for more than 45 days, unless exceptional circumstances applied. The UT stated that the FTT had erred in law in deciding that moral obligations and obligations of conscience can qualify as exceptional circumstances.

The UT were particularly critical of the taxpayer’s vague record keeping and inability to account for her movements for the five days in question. This was compounded by the fact that the taxpayer did not stay overnight with her sister for all five days, but instead spent some nights at her family home with her husband who was still living just seven miles away from her sister.

The taxpayer’s case was further weakened by the fact that her sister had two very good friends that would check on her several times a day. Her brother also lived 20 miles away and the taxpayer admitted that he had promised to keep an eye on their sister.

Conditions required to satisfy exceptional circumstances

The key finding of this case was that a moral obligation and obligations of conscience in themselves do not constitute exceptional circumstances. Paragraph 22(4) contains five conditions, all of which must be satisfied for each of the days for which the taxpayer is claiming the exceptional circumstances exemption for it to apply:

• the circumstances were exceptional;
• the circumstances were beyond the person’s control;
• the person would not be present in the UK at the end of the day but for those circumstances;
• the circumstances prevented the person from leaving the UK; and
• the person intended to leave the UK as soon as those circumstances permitted.

It was concluded that serious illness and death are not exceptional nor out of the ordinary. A moral obligation does not alter this position. The sister’s alcoholism and depression themselves and their consequences for family members, including children, are not uncommon or unusual. The definition of exceptional is not defined and its ordinary meaning is applied. The circumstances themselves did not stop the taxpayer from leaving the UK as alternative arrangements for the care of her sister and her children could have been made without the requirement for her to remain in the UK. Additionally, it was found that the taxpayer had made no care arrangements following either visit.

Take away points

It is crucial that individuals keep clear records of their day counting and their activities during their time in the UK, particularly if they are seeking to rely on the exceptional circumstances exemption for any of those days. Whilst an individual is not expected to produce “an itemised timeline for each day”, sufficient evidence must be provided to show that all five parts of the statutory test apply on each individual day.

HMRC give examples under paragraph 22(5) of what could be construed as exceptional circumstances, such as war, civil unrest or natural disasters or a sudden life-threatening illness or injury. The sister’s alcoholism and depression were not sudden and were documented as far back as 1996.

HMRC provide very prescriptive examples in their guidance such as explosion on a ship resulting in specialist hospital care or a parent staying in the UK to look after their minor child in hospital following a serious accident. The exceptional circumstances are limited to 60 days and if those circumstances surpass this amount, they will be included in the individual’s day count for the purposes of the SRT.

For exceptional circumstances to apply, it must be something out of the ordinary, outside of the individual’s control and but for those circumstances, they would not have been in the UK. Whilst certain circumstances may be upsetting and infer a moral obligation (such as the serious illness of a family member), this does not qualify these circumstances to become exceptional just by their underlying nature coupled with a moral obligation. Whilst quite blunt, the UT stipulated that serious illness is commonplace and death is universal and could not be considered exceptional circumstances.

If you have any questions regarding the SRT or what is considered exceptional circumstances, please contact a member of the Tax, Trusts & Succession team.

Michelmores has announced the appointment of Cheryl Brady as a Partner in its Transactional Real Estate team.

Cheryl qualified as a real estate lawyer in 2006 and has extensive experience within both commercial and residential property. Prior to focusing on property development in 2012, she dealt with a wide variety of commercial property transactions.

Since 2012, Cheryl has advised on property transactions which involve potential development of the land. She has acted for both landowners, developers and promoters through the complete life cycle of a development including site assembly, planning, site acquisition, funding, site setup of the development and most importantly the onward sales to housing associations, other developers, commercial purchasers and plot purchasers.

Cheryl currently focuses on acquisition of land for development for housebuilders. She is involved in a variety of sites including large mixed-use development sites and small sites often with complex issues to be addressed.

Commenting on the move, Cheryl said:

“I am excited to be joining Michelmores at a period of continued growth for the Firm. The Firm’s Real Estate group continues to go from strength-to-strength and I am looking forward to being part of its continued success over the coming years.”

Emma Honey, Head of the Real Estate group at Michelmores, comments:

“We are very pleased to welcome Cheryl to the Real Estate group. With Cheryl’s considerable expertise, I look forward to seeing our Real Estate department continue to progress and I congratulate her on this important appointment at the firm.”

Cheryl joined the Firm as a Partner in September 2023.

Following the governments proposal to scrap Nutrient Neutrality, it has been announced that the proposed amendments to the Levelling Up and Regeneration Bill (LURB) have been blocked by the House of Lords.

Concerns were raised about the impact these amendments would have on water pollution, further worsening an already significant issue. Additionally, the proposed amendments meant that Local Planning Authorities (LPA) would have to assume that wastewater from developments would NOT adversely affect a habitat site. This assumption would have applied to a vast range of planning decisions.

Baroness Hayman of Ullock, opposing this amendment, states, “we do not agree with the powers being introduced” by the proposed amendments which would mean, “abandoning legal protections for the nation’s most precious and sensitive habitats” as being the only means of meeting housing requirements. Adding that the way in which the Government have introduced these amendments is “entirely inappropriate.” Being proposed at the very last minute together with “excessive regulatory powers” and with no public consultation, which the Baroness states is, “frankly, astounding.”[1] Baroness Scott of Bybrook on the other hand, disagreeing with Baroness Hayman’s proposal, states that, ” What is being proposed instead amounts to dithering and delay—and adds confusion rather than clarity.” Instead, the Baroness emphasises that the additional investments being put into Natural England’s nutrient mitigation schemes, such as, nutrient credits, will ensure that any additional nutrient flow from new housing will be offset.

The Government defeat is not a surprise given that they have tried to significantly amend a Bill at the final reading stage. That was ambitious on any analysis but even more so when the proposed changes included powers to amend any EU or domestic environmental legislation. The Government could try and introduce a standalone Bill to deal with nutrient neutrality, but the question is whether they have enough Parliamentary time to do so before the next general election. There is no ability for the House of Commons to further debate this issue as part of the process of passing the LURB. The House of Commons can only debate amendments from the House of Lords and there were none as the amendment was rejected.

The result restores the status quo with a private market for phosphate and nitrate credits supplying diversified income for farmers and landowners. Existing deals for such credits will proceed although it is notable that all of the transactions that Michelmores are instructed on were unaffected by the Government’s announcement about the LURB amendments on 29 August. The private market for nutrient credits is developing well and will only become more efficient as time goes on. Market forces will establish realistic pricing. Developers and landowners need to collaborate in private deals against a backdrop of Governmental consistency and not uncertainty.

[1] https://hansard.parliament.uk/lords/2023-09-13/debates/4DD90393-6CD9-4F2D-A340-E99C96A8DF62/Levelling-UpAndRegenerationBill

On 31 August 2023, the House of Commons’ Science, Innovation and Technology Committee published its interim report on the governance of AI which covers many things including the question of AI and copyright. Unfortunately, the report does not reach any firm conclusions and so, until legislation is adopted, we will have to see if the UK Intellectual Property Office’s development of a voluntary code of practice on copyright and AI offers any answers.

So what is the current position regarding copyright and AI?  In this article, we are going to discuss, with the increasing use of AI:

• why AI created works are not protected by copyright in the UK;
• what it would take for such works to be protected by copyright; and
• how the meanings of “original” and “author” will need to change if AI created works are to be protected by copyright.

Having spent quite some time writing this article to answer the above questions (with no help from AI), we assert our copyright and take full responsibility for its content.

This article describes why the requisite intellectual creativity is lacking in an AI generated work and why we hope that, in order to protect the intellectual creativity human authors have put in, copyright will not be afforded to AI generated works for quite some time. Not least because if you asked AI to answer these questions, AI is not liable for its output if it gets the answers wrong!

It is important to remind everyone that the primary purpose of copyright law is to reward an author for the creation of an original work. It took more than 250 years after the development of the Gutenberg printing press for legislators to recognise this with a copyright statute[1] and so perhaps we should not be too critical of the government for taking its time over this question in the context of AI.

Meaning of author

Section 9 of the Copyright, Designs and Patents Act 1988 (the Act) defines “author”, in relation to a work, as the person who creates it; and, according to section 154 of the Act, a work qualifies for copyright protection if the author was at the material time a qualifying person[2]. AI does not satisfy the definition of qualifying person. There is some debate as to whether AI generated works fall within the definition of “a literary, dramatic, musical or artistic work which is computer-generated” so as to make “the author” “the person by whom the arrangements necessary for the creation of the work are undertaken” but the general view is that it does not.

Therefore, in order to qualify the person making the arrangements for the AI generated work as the “author”, this definition will need to make clear that it covers AI generated works.

Meaning of original

Another problem is that at present, AI generated works are not “original”.

The Act does not contain a definition of “original work”, but case law has developed the precedent that works are original when they are independently created as the “author’s own intellectual creation”. Independent creation simply means that you create the work yourself, without compromising it with, or copying it from, another pre-existing work.

AI produced works fail this definition. This is because AI works are neither:

• independently created because the AI is mining pre-existing sources; nor
• are they the result of the author’s own intellectual creation because the output is based on probability and algorithms not intellect. What has to be considered is the fact that AI is performing an essentially mechanical function dictated by its software engineers.

If we set AI generated works against this background, it is more understandable why the immediate output resulting from an input into ChatGPT is not afforded copyright protection.

What about copyright infringement?

Further, it is important to be mindful that in creating a work, AI uses all the available sources out there and if the content used by AI is already protected by copyright, then there is the risk that the work created using AI, will amount to copyright infringement.

Does this mean that an AI generated work is automatically an infringement of copyright?  Not necessarily as a copyright owner needs to show that a “substantial part” has been reproduced. This could be very difficult to establish unless the AI was operating on a very narrow body of underlying texts.

Legislative changes

Therefore, the legislation will need to address the definition of “original” to include AI works derived from the works of others. In addressing this issue, the legislation will need to consider whether:

• compensation should be paid to the copyright owners of the works on which the AI works are based; and
• the test for copyright infringement needs to have a “fair dealing” defence which allows the creation of AI works based on AI’s application to an existing body of copyright protected works.

So what if you ask AI to write a story for you today? Does the above analysis mean that copyright will never subsist?

Not necessarily.

The answer will very much depend on the degree of intellectual creativity that the real person, who wants to qualify as the author, applies to the initial draft of the AI generated work. The more intellectual creativity applied, the more the human author’s skill, labour and judgment is likely to qualify their contribution as “original” and create a protectable copyright work.

A moral question?

There is also a moral question (and one of moral rights).

Take for example a screenplay entirely generated by ChatGPT (and without further amendment or editing). Would it be fair and in line with copyright principles, if the “screenwriter” who asked AI to create the draft, to sit back and relax whilst still enjoying a nomination for “the best screenplay” award based on an original copyright work for which they claim the paternity right?

As things stand now, the answers are no, it wouldn’t be fair; and no, it wouldn’t be in line with copyright principles to claim to be the author.

Idea vs expression of the idea

It is also important, however, to remember that copyright does not protect ideas, but protects the expression of the idea. Therefore, if in creating a plot for you, AI suggests a story about  a young wizard, who goes to a school of magic, and his purpose is to fight a bad guy, then as long as that young wizard is not called Harry Potter, the school of magic is not called Hogwarts and the bad guy is not called Voldemort, you might still be able to create a successful and original work without infringing another author’s copyright[3].

Conclusions and takeaway points

The copyright implications for works generated by AI are still unclear and rapidly developing. The Government is aware that this is an area that needs to be watched closely and which needs to be regulated properly and soon[4]. Until then, no matter how complex the area is, the basic principles of copyright will apply.

Therefore, to avoid facing infringement proceedings and/or disappointments, use common sense and ask yourself: “am I proud of my work?”, “is this really my work?”, “is this work original and creative?”, if the answers are “no”, then think again. Where an author creates a work without reference to any existing subject matter, it will be rare that it does not attract copyright protection.  Even if you use AI as your starting point, if you make that work your own, then copyright will be available to you.

Please contact Iain Connor and Lorenza Picciano, who specialise in intellectual property disputes, if you have any questions about intellectual property or brand management.

[1] Statute of Anne 1709

[2] S.154 of the Act

[3] Please, refer to our previous article here: When your copyright claim disappears in a puff of smoke … ! to understand more on this point.

[4] The government’s response to the House of Commons Science, Innovation and Technology Committee’s interim report is due by 31 October 2023, so it will be interesting to see whether the momentum continues over the next couple of months.

Michelmores has advised Cornwall’s St Austell Brewery on the acquisition of The Bath Pub Company, growing its 184-strong West Country pub estate.

The deal, which completed for an undisclosed amount, sees the independent company take over the operation of four of Bath’s most well-known pubs.

The Locksbrook Inn, The Moorfields and The Hare & Hounds bring the total number of sites in St Austell Brewery’s managed pub estate to 43. The Marlborough Tavern is now part of the company’s leased and tenanted estate and will continue to be operated by Justin Sleath, The Bath Pub Company’s co-founder and commercial director.

The acquisition of The Bath Pub Company further strengthens St Austell Brewery’s presence in and around Bath where it already owns four tenanted pubs – The Griffin, The Swan, The Hop Pole and The Salamander. Michelmores also helped the company acquire Bath Ales in 2017, a deal which included a brewery in Warmley and a local distribution network.

The Michelmores team advising St Austell on the transaction was led by Partners Richard Cobb and Richard Walford, with support from partners Karen Williams, Cathy Bryant and David Thompson and lawyers Victoria Miller, Jonny Lane, Siobhan Murphy, Gruff Cartwright, Zilah Nelson and Justin Barrow.

Richard Cobb comments: “We always enjoy working with St Austell brewery and are pleased to have helped the team secure The Bath Pub Company. This is an exciting acquisition of a high quality and well-located portfolio of pubs with substantial investment and gives four more good reasons to be in Bath!”

Kevin Georgel, Chief Executive, St Austell Brewery said: “We’re extremely proud to become the new owners of such a high-quality and well-respected pub company in Bath.

“We are fully committed to developing the quality and size of our pub estate. As such, we are investing significantly in our existing pubs, whilst also acquiring new sites and businesses such as The Bath Pub Company. The Locksbrook Inn, The Moorfields, The Hare & Hounds and The Marlborough Tavern all have a strong alignment with our long-term plans and are a perfect fit for us, in terms of market and consumer positioning.”

Michelmores’ award-winning Corporate team of 25 specialist lawyers advises clients across the UK, US, EMEA and beyond – on capital markets, mergers and acquisitions, management buyouts, impact investing, energy projects, microfinance initiatives and more.

Find out more about our Corporate team here.

Nutrient neutrality has been in the political crosshairs since its inception and I have written previously about the fragility of the marketplace Nutrient neutrality: Are developments with outline consent caught by the requirements?.

Developers see it as a time consuming barrier to new houses which don’t contribute as much nutrient pollution as other sectors. Environmentalists regard it as a useful tool which ensures that new development does not contribute to an already significant pollution problem. Offering a solution are those involved in the private market for nutrient credits generated by taking land out of agricultural production.

Developers assert that 100,000 new homes are currently being held up by the requirement for nutrient neutrality and this is borne out by the experience of our residential development team. Others have argued that this statistic does not take account of the maturity of the private nutrient credit market which was becoming established.

In support of those assertions, a number of stakeholders, including the National Trust, CPRE and the Woodland Trust, co-signed an open letter to Rishi Sunak on 20 July 2023 which explained that over 70,000 homes have existing mitigation measures in place or in the pipeline. What will happen to those private credit deals which are agreed or in the process of being agreed, or the schemes created by numerous local planning authorities?

Notwithstanding that, the concerns of the developers seem to have been listened to as amendments to the Levelling Up and Regeneration Bill (“LURB”) have been proposed which will remove nutrient neutrality as a planning consideration if the Bill is approved by Parliament. The report stage of the Bill is to be concluded in the next fortnight, during which these amendments will be debated and voted upon.

A number of questions arise:

1. What is changing?
2. How will those changes affect existing private credit deals?
3. What should developers and landowners do in the light of this announcement?
4. How will the problem of nutrient pollution now be tackled?

What is changing?

The Government amendments to the LURB tabled on 29 August will change the Habitats Regulations so that nutrient pollution from housing developments will no longer be part of the planning process.

New clauses will be inserted into the Conservation of Habitats and Species Regulations 2017 (“Habitats Regulations“) which will require a local planning authority to assume that nutrients in urban waste water will not affect any relevant protected site. Those changes will only apply to developments where the wastewater is treated by a wastewater treatment works or private treatment system regulated by an Environmental Permit.

These changes are made possible by a new clause in the LURB at 159A which gives the Secretary of State wide powers to publish regulations which “may make any provision the Secretary of State considers appropriate” about the operation of any “relevant enactment.” That definition includes any domestic legislation or retained direct EU legislation relating to the environment so the winds of change may strengthen.

As a result, an appropriate assessment under the Council Directive 92/43/EEC on the Conservation of Natural Habitats and Wild Flora and Fauna (“EU Habitats Directive”) will no longer be required.

The amendments will have to reverse the current position, established in C G Fry v Secretary of State [2023], where the High Court determined that the relevant provisions of the Habitats Directive remain part of UK law due to the European Union (Withdrawal) Act 2018 (“Withdrawal Act 2018“).

How will those changes affect existing private credit deals?

The LURB amendments include a provision that the amended section 159 will come into force on “such day as the Secretary of State may by regulations appoint.” This is in contrast to the vast majority of the remainder of the LURB which becomes law on the day of Royal Assent or within 2 months of that date. A reasonable punt at a date of Royal Assent might be sometime in October which would suggest that we won’t see any regulations implementing these changes until some time in 2024 at the earliest.

Given the likely delay in publication of the necessary implementing regulations, the parties to any existing private credit deals may well decide just to carry on as the opportunity cost of trying to extract themselves may be too great. The likelihood is that those already involved in deals have developments which are ripe for commencement and as such may as well proceed. This may be because the cost of the credits has already been factored in or that reneging on deals will bring forth legal costs, loss of deposits and potential reputational issues.

In addition, the reality is that although the medium for compliance may have changed the cost will remain. The GOV.UK announcement 100,000 more homes to be built via reform of defective EU laws – GOV.UK (www.gov.uk) confirms that it “intends to work with the house building industry to ensure that larger developers make an appropriate and fair contribution to this scheme….” which perhaps indicates that this announcement is better news for the small and medium sized developers. In any event, there will be a cost associated with nutrient pollution and whether that is met through a private credit, Nutrient Mitigation Scheme Credit or through funding capital works as part of a Protected Site Strategy will be determined by the cost of each option.

The Michelmores Natural Capital Team have completed numerous private credit deals and those were already being influenced by the auction results of the Nutrient Mitigation Scheme Auctions in the Tees Valley Catchment. Those realised prices form a floor in the market and although that market is catchment specific once the Nutrient Mitigation Scheme is rolled out more widely (pump primed by £280 million of promised funding) then that pricing effect will similarly ripple outwards.

It follows that the market for private credits will likely remain providing they are priced competitively alongside the other mitigation options. They remain a useful turn key solution for developers who are naturally focussed on the business of building houses. It is then a question of land management economics as to whether it is worth landowners entering into such schemes or catchment approaches and nature based solutions .

What should developers and landowners do in the light of this announcement?

This is a difficult question to answer at this early stage. We still have to wait for the comments of the House of Lords and any possible changes they propose. The immediate effect of the announcement is to increase the levels of uncertainty and that will not change until we see the secondary legislation that is authorised by the LURB.

I have been asked whether such a change in policy can be challenged. There is quite a lot in that question but essentially the Government will be seeking to change domestic legislation to ensure that nutrient neutrality is no longer part of the planning process.

Brexit has made these changes possible as nutrient neutrality is a creation of European law with the Habitats Regulations transposing the requirements of the EU Habitats Directive.

Since Brexit, domestic legislation derived from EU law, such as the Habitats Regulations, continues to have effect pursuant to section 2(1) of the Withdrawal Act 2018. Similarly, the pre-Brexit case law of the CJEU lives on in relation to the interpretation of EU law.

The LURB amendments will amend the Habitats Regulations and will have to get through Parliament to become law. If that process is completed lawfully then the Government will only be susceptible to Judicial Review if they make decisions in respect of that new legislation which are wrong or unreasonable.

During this period of continued uncertainty, developers (and landowners with their own development aspirations) who wish to bring forward projects may conclude that securing private credit deals is the best option. This is an alternative to speculating on political outcomes and whether or not the new regime is a more cost effective solution than the existing one.

The Government suggests a variety of alternative strategies to tackle the problem, including increased funding for Natural England’s Nutrient Mitigation Scheme and the use of catchment scale schemes and Protected Site Strategies, meaning there will still be an important role for landowners in providing nutrient mitigation solutions if the amendments become law.

How will the problem of nutrient pollution now be tackled?

The Government has announced a package of measures to combat the problem of nutrient pollution including the expansion of the Nutrient Mitigation Scheme as set out above and the implementation of Protected Site Strategies.

Water companies are also being encouraged to actively consider nature based solutions to treat wastewater as well as the statutory 2030 upgrades. A catchment permitting approach will be allowed so water companies will be judged on the collective performance of all wastewater treatment plants. This would, in theory, allow some wastewater treatment plants to discharge at levels above the accepted limit if other plants were over achieving.

The GOV.UK announcement refers to £225 million of funding to reduce runoff from agriculture which will be policed by 4,000 inspections each year to ensure water pollution is minimised.

The LURB amendments will have no effect beyond 2030 which is the deadline for the upgrading of wastewater treatment works in designated catchment areas. My back of an envelope calculations estimate that if those upgrades are achieved then the requirement for nutrient credits would roughly halve. The pressure is on the water industry to improve infrastructure but funding has also been made available to enable other sectors to reduce their impact on protected sites.

The Government have previously stated that they intend to “crowd in” private schemes rather than crowd them out and allowing the private credit market to be one of the mitigation solutions would achieve that goal.

Whilst recent developments in Artificial Intelligence (‘AI’) services might have some of us fearing increasingly harmful cyber-attacks, evermore convincing deepfakes and even seeing Robert Patrick’s T-1000 chasing after us in our rearview mirrors, we cannot ignore that AI is already an important part of our day-to-day lives.

First invented in the 1930s by Georges Artsrouni, AI has gone through numerous evolutions and is still traditionally focused upon detecting patterns, automation and generating insights. It is currently employed in the workplace – pun very much intended – to undertake tasks such as filtering spam, automated CV screening, task allocation and performance management. The use of this type of AI has long been widely accepted within the workplace.

Generative Artificial Intelligence (‘GenAI’) is a type of AI, which learns from existing data patterns to produce new types of content, such as text, imagery, videos, audio and synthetic data. This has been a common part of our day to day lives since the introduction of basic chatbots in the 1960s. However, with the introduction of OpenAI’s chatbot ChatGPT in November 2022 and more recently Microsoft’s Co-Pilot in March 2023, GenAI has become far more advanced and can be used to solve complex problems, draft articles in seconds – unfortunately for me, not this one – and even prepare detailed and entertaining speeches and presentations. It’s also become incredibly user friendly and, without any sign-up costs, entirely accessible to the average person. It’s therefore hardly surprising, that more and more people are using it. And that’s the rub!

GenAI sounds awesome, so why not use it in the workplace?

According to the latest available data, ChatGPT currently has over 100 million users. And the website generated 1.6 billion visits in June 2023. It’s not hard to see why it is so popular: ChatGPT generates responses, which are quick, contextually relevant and ‘human like’. However, there are a number of limitations with its function, which means that relying on its responses can be inherently risky. ChatGPT learns from the data inputted by its users, which it then uses to inform other users. This means that if users input sensitive, fabricated, biased or, indeed, malicious data, this can then be presented by ChatGPT to other users as fact. Now you can start to see why this would make employers and, well, any of us a little nervous…

Our recent article on AI discusses this and the possible ramifications for human roles within businesses more broadly.

Hang on, surely the government is going to legislate so that users and employers are protected, right?

Whilst there is no doubt that the use of GenAI can increase productivity and be an effective tool to aid employees in their roles, appropriate safeguards must be put in place to manage risk and protect businesses.

You may recall that in March 2023, the UK government’s White Paper confirmed that the UK did not intend to introduce specific legislation nor a single governing body to regulate AI; instead it would support existing regulators to regulate AI in their sector. Following on from this, the House of Commons library published a paper on 11 August 2023 on AI and employment law, which assesses how AI is currently (and will in the future be) used at work, alongside the current legislation and policy developments.

So we need to take steps to safeguard our businesses

How your employees use GenAI is likely to depend on the sector in which your organisation operates, and the type of work it carries out. As a medium-term option, we would encourage businesses to undertake a review of (1) how the people in their organisation are currently using Chat GPT and other GenAI and (2) how these tools might be used by their organisation/employees in the future, so that they can tailor their safeguards accordingly.

What’s the rush?

This, however, overlooks the immediate issue…employees are using ChatGPT and other GenAI now! With the staggering figures quoted above, it stands to reason that many of these users will be using ChatGPT et al for work related purposes. Therefore, employers need to work fast and get a GenAI policy in place as quickly as possible.

With GenAI growing in competence every day and user numbers similarly building, smart employers should be getting a basic policy in place immediately and then looking to finesse and tailor that policy to their business and sector needs over the coming weeks. Failure to do, puts businesses at risk of their employees sharing sensitive company and client data via ChatGPT and using it to obtain information and documents, which may well contain fabricated, biased and/or malicious data.

GenAI policy

Okay, well, what should this basic policy include?

When looking to introduce such a policy, consideration should be given to the following:

• Level of prohibition: will you prohibit the use of GenAI completely or will you put limitations on who can use it and/or what it can be used for?
• Guidelines for its use: if you are going to allow staff to use tools like ChatGPT for certain tasks, you should make sure the permitted functions are expressly listed so there can be no confusion.
• Set parameters for its use: general guidelines on how to use GenAI should be expressly set out in the policy, for example:
• It should only ever be used as a starting point, not as a finished product.
• All content must be proofread and checked for factual accuracy by a human with appropriate expertise before it is used.
• Confidential information and personal data should never be divulged. Even things like your company name and other identifying features should not be disclosed when using GenAI.
• Highlighting its limitations: your staff need to be aware of GenAI’s inherent restrictions. For example, ChatGPT is based on data stored in its bank and so if that data is incomplete, inaccurate or biased (even discriminatory), it means the responses produced will be the same. Also be mindful that the bank is outdated and doesn’t have the most recent information available.
• Designated team or person: will you have a steering group or designated individual who will oversee the company’s approach to GenAI?
• Transparency: consider how you will go about ensuring that AI generated content is clearly identified as such – both internally to other employees and externally to clients.
• Data privacy and confidential information: as highlighted above, it is essential that no confidential information or personal data is shared with GenAI tools. For example, ChatGPT has no obligation to keep this information private and instead can use the information to improve and develop its systems.
• Interaction with other policies: you will need to consider how your ChatGPT policy will interact with other relevant policies, such as data protection, IT and communications, privacy, recruitment (if you are using ChatGPT/AI to make recruitment decisions) your disciplinary policy and even your grievance policy if GenAI is misused or there’s a breach of your GenAI policy.

Additional points to consider

As well as having an effective policy, running training sessions and an awareness campaign should help embed expectations and encourage employee buy-in. Those of us that experienced the internet and then the social media revolutions in the workplace will know all too well that this is an incredibly fast moving area, and your policy will have to be regularly monitored and updated to ensure it’s up to date and manages risk appropriately.

If you’d like help drafting a GenAI policy, or if you have any other AI-related employment or immigration queries, please do not hesitate to contact Lynsey Blyth.