In today’s Media News Ireland landscape, the biggest hiring story may not be who applied for the job, but who wrote the application. A revealing recruitment test at international law firm Addleshaw Goddard suggests AI tools may rank CVs more favourably when those CVs appear to have been shaped by AI themselves.
That possibility matters far beyond one employer. As recruiters across Ireland and beyond face a flood of applications, artificial intelligence is increasingly being used to speed up screening. But if those systems show a built-in preference for polished, AI-assisted applications, the hiring process could tilt in ways many companies have not fully considered.
Media News Ireland: The New Recruitment Dilemma
The case emerged after Addleshaw Goddard advertised an IT role within its HR function and received 178 applications, a sharp jump from the kind of volume that might have landed just a year earlier. The rise reflects a clear trend in Media News and workplace reporting: AI has made it dramatically easier for candidates to apply for jobs at scale.
Faced with the challenge of reviewing so many applicants quickly, the firm ran a practical experiment. Its HR team used an internal Microsoft Copilot-based agent, gave it the role criteria and asked it to score a shortlisted group of candidates. Human reviewers did the same.
The result was striking. The AI’s preferred candidates were not the same as the humans’ top picks.
According to the firm’s HR leadership, the AI appeared to favour highly polished applications that presented big achievements, rapid career wins and impressive project claims, even where the experience looked thin. In editorial terms, the message is simple: the machine may have been seduced by the shine.
Why AI May Prefer AI-Written CVs
There are several plausible explanations, and they are now becoming central to News Ireland coverage of the future of work.
1. Better keyword matching
AI tools help job seekers tailor CVs closely to a job description. That means applications may mirror the language, structure and priorities of the role more precisely than a human-written CV would.
2. AI signals mistaken for AI skill
In this case, experience with AI itself was one of the desired capabilities. That may have led the assessment tool to treat signs of AI-assisted writing as indirect evidence of AI literacy.
3. Self-preferencing by the model
This is the most unsettling possibility. If an AI model inherently rates content that resembles machine-generated output more highly, then applicants using similar tools could gain an invisible edge over candidates who write their own CVs.
That concern is not just theoretical. Research cited alongside the case found that applicants using the same large language model as the recruiter’s tool were significantly more likely to be shortlisted. Depending on the role, that advantage ranged from 23 per cent to 60 per cent.
What This Means for Employers
For hiring teams, this is more than an interesting experiment. It is a warning shot, and one that deserves attention across Agency News Ireland and business reporting circles.
If AI tools are allowed to screen applications without careful oversight, employers risk:
- Rewarding style over substance
- Missing strong candidates whose CVs are less polished
- Introducing hidden bias into recruitment decisions
- Creating unfair advantages for candidates using specific AI tools
- Accelerating a hiring arms race that overwhelms recruiters
The Addleshaw Goddard example also highlights a basic operational issue. The faster companies move to automate recruitment, the more important it becomes to understand what the technology is actually rewarding.
What This Means for Job Seekers
For candidates, this story lands at the intersection of Corporate News Ireland and career strategy. Many already assume they are applying into systems where software scans every line before a human ever sees the file. That belief has fuelled a surge in AI-assisted applications.
The risk now is a two-speed market:
- Candidates using AI to sharpen, tailor and optimise their CVs
- Candidates relying solely on traditional writing and formatting
If the screening tools themselves favour AI-shaped language, then applicants who avoid AI could be disadvantaged before they even reach interview stage.
Still, there is an important caveat. Human recruiters in the firm’s trial did not select the same top candidate as the machine. In the end, the successful hire was chosen by people, not by the AI’s highest ranking.
That detail matters. It shows AI can assist the process, but should not define it.
A Smarter Way to Use AI in Hiring
There are practical ways to reduce the risk of AI-on-AI bias, a point that deserves wider attention in every serious Media Digest on recruitment technology.
Researchers suggest two relatively simple fixes:
- Instruct the model to ignore the origin of the CV and focus strictly on relevant skills, experience and evidence.
- Use multiple assessment models rather than relying on a single tool, reducing the influence of one system’s built-in preference.
Companies should also consider broader safeguards:
- Keep humans involved in final shortlisting decisions
- Audit AI scoring patterns regularly
- Test whether certain writing styles are being overrated
- Assess real capability through interviews and work samples
In short, AI can speed up hiring, but speed is not the same as judgment.
The Bigger Picture for Media News Ireland
This recruitment story is really about trust. Employers want efficiency. Candidates want fairness. AI promises both, but only if it is designed and supervised properly.
For readers following Media News Ireland, the takeaway is clear: the future of hiring will not simply be humans versus machines. It may increasingly become a contest between one machine’s preferences and another machine’s output.
That should concern every employer using automated screening and every applicant wondering how much algorithmic polish is now required to compete. The technology is useful, but it is not neutral by default. As this case shows, human judgment remains the essential backstop.
In the end, the lesson for Media News Ireland readers is a sharp one: AI may help write the modern CV, but companies must make sure it does not quietly decide that only AI-written candidates deserve to rise to the top.
