Experts from William Fry take a closer look at AI and automation’s effects on the working world, particularly in terms of making hiring decisions.
Automation in employment is nothing new. From water-powered flour mills to Excel spreadsheet formulae, automation in some form has been around almost as long as the concept of employment. However, what is new is the rapid expansion of AI technologies and their ability to have profound and immediate effects on the workplace.
Many people, no doubt fuelled by occasionally sensationalist headlines, fear that AI bots could replace all of us in our employment in the coming months. A recent report found AI could replace up to 300m jobs worldwide. While AI is progressing at a startling rate, and the law is most certainly playing catch-up, this level of automation is not expected shortly.
The reality is that AI is already present in almost every workplace, at least in some form – for example the smart engines that predict where to file your emails or correct your typos. We can see that AI has therefore quietly been present in the workplace for some time. There will be some changes to the way we work based on recent AI advances and repetitive work suited to automation is clearly at risk.
However, the AI uber-machine will not be able to make the decision about which jobs should go and which jobs will stay. Established Irish employment law, data protection law and upcoming EU legislation on AI will ensure that any decisions relating to jobs and AI will still be made by humans.
The AI Act
The protection against automated decision-making is an existing concept that legislation will likely expand on in the future, such as the draft EU regulation on AI, the AI Act. The AI Act is a draft piece of EU legislation expected to come into law in late 2025 and recently made headlines due to reaching some landmark points in its journey through the EU legislative system.
The AI Act is expected to do to AI activity what the GDPR did to data processing – which is, put in place at least a rudimentary structure by which the law can contextualise and interact with an area of technology which is shifting constantly.
The AI Act classes AI systems which deal with employment as ‘high risk’. High-risk systems will be subject to further scrutiny and control: regular formal risk assessments, data processing impact assessments and onerous record-keeping requirements.
Annex III of the AI Act lists the specific activities where the use of AI is to be considered high risk and one of those activities is the use of AI in an employment context, such as performance monitoring or CV filtering.
While there is more debate to come in relation to the AI Act, the inclusion of AI in employment settings has not been the subject of much, if any controversy, and while we await the final text, it is not looking like AI in the employment context will change drastically.
This means that any organisations using AI in this way, or creating AI systems to be used in this way, will have strict regulatory obligations under the AI Act, such as record keeping, transparency, risk management and data governance.
GDPR
Article 22 of the GDPR relates to automated individual decision-making, including profiling. It says that individuals have the right not to be subject to a decision based solely on automated processing, including profiling, which produces legal effects concerning him or her, or similarly significantly affects him or her.
When applied to the context of AI and employment, it’s highly relevant. Many companies use AI systems and automated processes for decision-making in several areas, including hiring, promotions, employee evaluation and even termination in some cases. These can include algorithms that filter CVs, automated interview systems, performance tracking software and even predictive models used for succession planning.
Under Article 22, individuals can opt out of such decisions based solely on automated decision-making. This means employers should provide some sort of human oversight or intervention in these processes. If an employer in the EU or dealing with EU citizen data uses AI for such decisions, they need to ensure that individuals are informed about the processing, its implications and their rights.
AI and employment equality
Some firms use (or are considering using) AI at the recruitment stage to initially screen and filter applicants. Already, it has been suggested that some AI programmes carry out this process without regard to relevant employment law protections or potential latent biases of the software programmers – for example, by filtering out certain language patterns which are more prevalent among different groups of people. This is something which lawmakers are very conscious of. Margrethe Vestager, EU competition commissioner, recently said: “I think the AI risks are more that people will be discriminated against, they will not be seen as who they are.”
In Ireland, equality is a cornerstone of our employment law framework – and it is well established that employment equality protections extend to the recruitment stage. Regardless of the progress of the AI Act, AI-led employment decisions have the potential to be unlawful under Irish law as is.
Already, individuals have the right not to be subject to automated decision-making under current GDPR rules. There are existing heightened data protection obligations under the GDPR. For example, providing information and certain types of data processing by AI systems may require a data protection impact assessment.
Employers should be alert to these potential impacts when they consider software. By way of example, if a piece of software was used to read CVs, it could consider ‘John’ to be the only correct spelling of a name – and disregard all applications from Jon, Seán, Jan or other similarly spelled or sounded names as ‘incorrect’. That is an obviously over-simplistic example, but now is the time for employers to honestly consider whether their employment systems as a whole, and not just in respect of recruitment, are sufficiently robust.
While AI is a developing area and there may be a learning curve in its adoption and implementation in the workplace, equality is a foundational element in Irish employment law and no excuses for breaches of that established legal framework can reasonably be expected to be tolerated by the Workplace Relations Commission or the courts.
In short, AI is already present in the workplace and automating a huge number of tasks which were once performed by people (such as filing) or were never done by anyone (such as predictive autocorrect).
While automation will continue to grow as AI keeps developing, legal structures are being developed now which will allow our established legal systems to interpret and apply employment and equality law in this new context.
By Catherine O’ Flynn, Barry Scannell and Oisín O’Callaghan
Catherine O’Flynn is a partner and head of William Fry’s employment and benefits department. Barry Scannell is a senior solicitor and consultant in William Fry’s technology department specialising in artificial intelligence. Oisín O’Callaghan is an associate in William Fry’s employment and benefits department.
10 things you need to know direct to your inbox every weekday. Sign up for the Daily Brief, Silicon Republic’s digest of essential sci-tech news.
