The impact of artificial intelligence (AI) on jobs is the subject of intense debate. But it is also changing the wider workplace: remaking recruitment, providing new – and sometimes controversial – tools for employee monitoring and transforming the process of employees making complaints. Workplace relations are now inundated with vast amounts of AI-created information in a kind of arms race between employees and employers.
Whether this is a temporary transition towards a better use of this technology remains to be seen.
Recruitment
The ability of applicants to use AI tools to create CVs and application letters – and scrape jobs site for opportunities – had led to a huge increase in job applications.
The sometimes-tiresome task of putting together an application has been transformed to, at most, a few clicks and sometimes an automated process where AI will fire off your CV in response to job ads it feels you are qualified for.
READ MORE
Jobs that previously attracted a couple of hundred applicants can now get applications into the thousands, according to Trayc Keevans, global foreign direct investment director at recruiters Morgan McKinley.
She cites a recent ad for a business analyst position, which might in the past have received 100-150 applications but this time it attracted 2,000 plus. Applications are also arriving much more quickly, she says, typically with 24 to 48 hours, as candidates use AI to move quickly, providing challenges to the systems of recruiters.
In turn, bigger companies are often using AI tools to screen and shortlist applications based on various criteria – an area where care is needed not to leave themselves open to legal action on the basis of discrimination.
The European Union’s AI Act imposes new obligations, and new parts are due to come into force next month. This obliges employers to be open and transparent in the use of AI in the workplace and adds to existing anti-discrimination obligations under equality law.
The flood of applications is causing challenges for companies. Recruitment software company Greenhouse found in a survey of recruiters, hiring managers and candidates in the UK, Germany and Ireland that many are struggling to keep up with early-stage reviews of applications with 38 per cent saying that half their work is spent “filtering out junk, spam or completely unqualified applications”.
This trend, Keevans says, can lead to frustration on both sides, with employers facing costs to trawl through vast numbers of applicants, some not properly qualified, and applicants at time feeling their application is not being properly assessed – perhaps not even being seen by a “human” – and they are not getting past the first hurdle.
Proper human oversight is essential, she says, even if the volume of applications means the use of AI in initial screening is inevitable.
The Greenhouse survey found mixed confidence among recruiters and hiring managers in AI tools used in this context – the right mix between AI and human judgment can be hard to find.
More than half believed that AI made it easier for applicants to cheat – via fake credentials, for example, or exaggerating their experience, while nearly half say they have seen candidates read off AI-generated responses during a live online interview.
The survey of candidates, meanwhile, says that many are embedding “prompt interjections” in their CV, hidden instructions – text hidden in LinkedIn profiles or white on white text in applications documents, for example – designed to confuse or bypass AI models. In effect trying to “game” the AI system to get to the next stage.
[ Almost half of Irish workers ‘worry about AI taking their jobs’Opens in new window ]
Candidates report declining trust in the hiring process, with close to half saying their trust in the hiring process has fallen over the past year, according to the Greenhouse survey.
Views on AI-assisted screening interviews were mixed. Frustration is being added to the slowing general recruitment environment in many sectors and the slow processes companies are using to make sure they get the right people.
Keevans says that AI has reduced the “individuality” of CVs: “They all look the same.” Employers have to search for evidence of skills beyond the academic essentials – often shown by personal interests and experiences.
With the market for graduates tightening, Keevans says candidates need to be prepared to look across different sectors for employment – beyond their initial target – and realise that with AI taking on some more basic tasks, companies are putting more emphasis on finding rounded recruits with strong interpersonal and problem-solving skills as well as good academic credentials.
Performance monitoring
Some big companies use AI to monitor or review employee performance, with some tools now saying they can even “predict” future performance. This can be a legal minefield for employers, particularly where human involvement in key decisions cannot be demonstrated. As far back as 2022, three employees of Estée Lauder subsidiary MAC, who were told they were being made redundant, partly on the basis of an video interview conducted by AI, received an out-of-court settlement.
The EU AI act will impose new legal obligations on companies – in addition to existing equality and data protection law. This will outlaw indiscriminate monitoring of employees using AI and imposes new obligations in terms of any monitoring, requiring – among other things – employees to be informed that AI is being used and requiring human oversight.
Employee complaints
AI has led to a sharp rise in the number of employee complaints to their employer – and to cases before the Workplace Relations Commission (WRC), which saw a 44 per cent rise in complaint applications in 2025 compared to the previous year.
Not all of this can be attributed to AI. But it seems a fair part of it can; as with job applications, making a WRC complaint is now much easier using AI.
[ Rise of AI threatens about 110,000 Irish jobs, study findsOpens in new window ]
The size of complaints has also expanded hugely, with companies finding themselves “inundated with paper”, according to Jennifer Cashman, employment partner at RDJ solicitors, with single complaints sometimes running to hundreds of pages of AI-assisted text. In turn, Cashman says, this is increasing costs for employers, often facing multiple claims and required to assess all the lengthy documentation and prepare defences.
In some cases AI may help genuine applicants to file claims. But in others it can lead them off track. “AI wants to be helpful,” Cashman says, “but it can also be wrong.” In an October 2025 case taken by a Ryanair employee, the adjudication officer at the WRC noted that the submissions were rife with citations that were either irrelevant, misquoted or in some instances non-existent. The employee lost the discrimination case.
The Workplace Relations Commission and the Labour Court have both published guidance on the use of AI, warning that all sides have a responsibility to ensure the information quoted is correct and underlining in particular that AI can “invent” legal cases or take them completely out of context. Neither requires parties to a case to disclose their use of AI.
The Protected Disclosure Commissioner has also seen a big increase in complaints, noting in his most recent annual report that many are “employee grievances, interpersonal disputes or complaints from people who are unhappy about a matter rather than reports of wrongdoing within the meaning of the legislation”.
The flood of paper is making it harder for companies to deal with internal grievances and WRC cases, Cashman says, and increasing costs for employers. For complainants, it makes filing the initial case easier, but AI is not proving reliable in compiling supporting evidence. In the meantime, the system struggles to cope.
