Artificial intelligence is now embedded in many recruitment platforms. It can search for candidates, screen CVs, score assessments and recommend who should progress to an interview.
For fintech employers, these tools can make recruitment faster and more consistent. However, under the European Union’s AI Act, certain uses of AI in employment may be classified as high-risk because they can significantly influence a person’s access to work.
Fintech recruiters therefore need to understand not only whether their software uses AI, but what that AI actually does.
Which Recruitment Tools May Be High-Risk?
An AI tool is more likely to be considered high-risk when it materially influences recruitment or selection decisions.
Examples may include systems that:
- Rank candidates according to suitability
- Automatically filter or reject applications
- Search platforms and produce candidate shortlists
- Score written or spoken interview answers
- Evaluate personality, behaviour or competencies
- Perform risk-based background assessments
- Decide which applicants see particular job advertisements
The fact that a recruiter makes the final decision does not automatically remove the risk. If an AI-generated score or ranking is a major influence on who progresses, the system may still fall within the high-risk category.
Not Every Recruitment Tool Is Treated Equally
Some AI-enabled tools perform limited administrative tasks without evaluating candidates. These may be treated differently from systems that make or influence selection decisions.
For example, software used only to schedule interviews, organise CV information or send application acknowledgements may present less risk if it does not rank, profile or assess applicants.
The key question is not simply, “Does this platform use AI?” Recruiters should ask, “Does this system influence someone’s opportunity to obtain employment?” That distinction should form the basis of every internal review.
Review Your Recruitment Technology
Many employers rely on applicant tracking systems without fully understanding the features operating behind them. A platform may have introduced AI-powered ranking, matching or assessment functions during a routine software update.
Create an inventory of every tool used across recruitment. Record what information each system collects, what outputs it produces and how those outputs affect decisions.
Fintech employers should also ask vendors direct questions:
- Which features use artificial intelligence?
- How was the system trained and tested?
- Does it rank or eliminate candidates?
- Can recruiters override its recommendations?
- How does the provider monitor discriminatory outcomes?
- What records and technical documentation are available?
- Where is candidate data processed and retained?
Marketing claims about being “fair,” “objective” or “compliant” should not replace proper evidence.
Keep Meaningful Human Oversight
Human oversight must involve more than approving whatever the software recommends. Recruiters should understand the system’s limitations and be able to question its output.
A qualified person should review unexpected rankings, investigate patterns and override recommendations when appropriate. Employers should avoid situations where recruiters become so dependent on automated scores that manual review is merely a formality.
Partnering with a specialist such as LibertyLoom Talent can also help fintech employers maintain informed human judgement when sourcing and evaluating candidates for specialised roles.
Test for Bias and Unequal Outcomes
An AI system may reproduce patterns hidden within historical recruitment data. This could disadvantage candidates based on gender, age, disability or another protected characteristic, even when the platform does not explicitly use that information.
Employers should monitor outcomes across different groups and investigate unexplained differences. Job requirements should also be reviewed to ensure the system is measuring qualities genuinely relevant to the position.
Testing should continue after deployment because candidate pools, job requirements and software models can change.
Give Candidates Clear Information
Transparency builds trust and helps candidates understand how their applications are handled. Employers should provide appropriate information when AI materially supports screening or assessment.
That explanation should be written in plain language. Candidates should not need technical expertise to understand that automated technology is being used or how it may affect the recruitment process.
Make AI Governance a Shared Responsibility
Compliance should not sit entirely with recruiters. HR, legal, compliance, information security and data protection teams should work together to evaluate hiring technology.
Fintech companies already operate within demanding regulatory environments. Applying the same governance discipline to recruitment AI can reduce legal, reputational and operational risk.
The EU AI Act does not mean employers must abandon useful technology. It means they need to understand it, document it and ensure people remain accountable for employment decisions.
#EUAIAct #FintechRecruitment #HiringTechnology #ResponsibleAI #HRCompliance #AIRecruitment #TalentAcquisition #FintechJobs #LibertyLoomTalent
