share on
AI is helping employers hire faster, but questions remain over bias, transparency, data privacy, and accountability. Umairah Nasir speaks to HR and legal experts about the risks organisations may be overlooking, and what it takes to use AI in recruitment responsibly.
Artificial intelligence (AI) is becoming an increasingly common part of the recruitment process. Employers are using AI-powered tools to screen CVs, shortlist candidates, schedule interviews, and identify potential hires more efficiently, helping recruiters manage growing application volumes while reducing administrative work.
Yet as AI takes on a bigger role in hiring decisions, concerns are growing over whether organisations fully understand the risks involved. While the technology promises speed and consistency, it also raises important questions around fairness, transparency, data privacy, and accountability, particularly when candidates may be assessed or filtered out before a human review takes place.
The issue has attracted global attention in recent months. Earlier this year, AI recruitment technology provider Eightfold AI faced a class action lawsuit in the US alleging that its technology generated candidate scores and profiles without applicants' knowledge or consent. While the case remains specific to one company, it has sparked wider discussions around how AI is used in recruitment and who should be held accountable when automated systems influence hiring outcomes.
As organisations continue to expand their use of AI in recruitment, the challenge is no longer whether to adopt the technology, but how to do so responsibly. To better understand the opportunities, risks, and safeguards involved, Umairah Nasir speaks to HR and legal experts about where employers should draw the line between automation and human decision-making, the legal risks they need to be aware of, and the steps they can take to ensure hiring remains fair, transparent, and accountable.
Read on to learn how the interviewees believe organisations can harness AI's benefits without compromising fairness, transparency, and compliance.
What HR leaders are seeing on the ground
Ong Eng Hui, Senior Director, Human Resources & Information Technology, Systems on Silicon Manufacturing Company (SSMC)
Q1 Is your organisation currently using AI in the recruitment process? If yes, what benefits or challenges have you seen so far? If not, are there plans to introduce AI tools in hiring in the future?
SSMC has been advancing digital transformation for several years, and more recently accelerated AI adoption under our connectivity, automation and intelligence roadmap. In HR, we are piloting AI in recruitment, focusing on areas where time is disproportionately spent — such as CV screening and initial shortlisting. These high-volume tasks often consume effort without significantly improving decision quality.
We expect measurable impact, including a 20% reduction in time-to-hire, 30% time savings in CV screening, and the ability to screen 100% of applications. Qualitatively, AI helps drive consistency in shortlisting and improves alignment between candidate profiles and job requirements.
Ultimately, AI allows recruiters to redirect effort toward more value-added, human-centric aspects of hiring.
Q2 What steps do you think companies should take to ensure AI hiring tools are fair, transparent, and do not unintentionally discriminate against candidates?
AI hiring tools must align with an organisation’s code of conduct, equal employment principles, and applicable labour regulations, such as Singapore’s Model AI Governance Framework or the EU AI Act. Importantly, AI should augment — not replace human judgement.
To mitigate bias and ensure fairness, organisations should implement strong human-in-theloop guardrails. This includes a final human review of all AI-selected candidates before progressing them, as well as regular audits of rejected profiles to validate accuracy and neutrality. Ultimately, hiring decisions must remain accountable to humans, who are better positioned to assess qualities like integrity, empathy, and leadership. These measures ensure AI supports a fair, transparent, and responsible recruitment process.
Q3 How important is human oversight in AI-driven recruitment, and where should HR draw the line between automation and human decision-making?
AI and human judgement should be seen as complementary. In recruitment, AI is highly effective for CV screening, identifying skill matches, and detecting patterns or potential bias. However, critical decisions such as final hiring choices, cultural fit, and evaluating potential versus experience must remain human-led.
This principle applies beyond recruitment. In workforce planning, AI supports forecasting and scenario analysis, while humans make strategic trade-offs around investment, workforce composition, and resilience.
The guiding principle is clear: AI excels at scale, speed, and pattern recognition, while humans provide context, judgement, and accountability. Especially in decisions involving fairness, employee impact, and long-term strategy, strong human oversight is essential.
Daniel Kusmanto, Senior Director Global Head People Data Process and System, ASM
Q1 Is your organisation currently using AI in the recruitment process? If yes, what benefits or challenges have you seen so far? If not, are there plans to introduce AI tools in hiring in the future?
Yes, we are leveraging AI in our recruitment process and intend to do more of it. I do see two-fold benefits, one is enriching our talent sourcing activity by opening up a new sourcing channel or simply by digging deeper and highlighting more leads from our existing talent pool, two is optimising our recruitment process, by scheduling interview, generating job description, note taking, etc.
The challenge is adoption and accuracy. The true value of it can only be realised when the solution is accurate and fully adopted. Sometimes this takes time to mature and we need to be able to invest time to embrace and train it. Starting small then scale up always helps.
Q2 What steps do you think companies should take to ensure AI hiring tools are fair, transparent, and do not unintentionally discriminate against candidates?
Governance is the key. This is not only for AI because human can also have bias. So having a good governance with the right metrics can help to ensure fair and transparent selection process.
Q3 How important is human oversight in AI-driven recruitment, and where should HR draw the line between automation and human decision-making?
Human brain can consider multiple factors, sense multiple cues hence I am not sure the current state of AI can already comprehend all of these. AI is proven to be able to collect relevant information effectively and come up with common practice fast. Hence, this could potentially be where we draw the line for now when things are more standard, operational, hard skills, exact, AI can kick in, but when assessment goes towards the soft skills, culture fit, or complex environment, we might still need to base it with human judgement.
AI can also be shaped as a co-pilot in bringing out relevant information while hiring managers decide.
At the end, governance and oversight are always a must to ensure fair and transparent selection process.
What legal experts say about AI hiring risks
Lin Yingxin, Director, KGP Legal
Q1 What are the biggest legal risks employers should be aware of when using AI tools in recruitment and hiring?
One significant risk pertains to discrimination. AI systems trained on historical hiring data may perpetuate existing biases, potentially resulting in claims of unlawful discrimination on grounds of race, gender, age, or disability. In Singapore, this may contravene the Tripartite Guidelines on Fair Employment Practices and/or the Workplace Fairness Act 2025 which is expected to come into operation soon.
There are also data protection obligations under the Personal Data Protection Act 2012 to consider, particularly around how candidate data is collected, used, and retained. Employers remain responsible for such legal risks even if they use AI tools.
Q2 Could companies face liability if AI systems unfairly filter out candidates or make biased hiring recommendations, even if the technology is provided by a third-party vendor?
Contracting with a third-party vendor does not insulate an employer from liability for discriminatory outcomes. The employer remains responsible under Singapore employment law for the decisions made in its recruitment process, regardless of the tool used to support them.
Third party vendors may share liability depending on contractual arrangements, but regulators and courts will look to the employer in the enforcement process. Employers should conduct due diligence on any AI tool they adopt and ensure contracts clearly address accountability, audit rights, and remediation obligations.
Q3 As AI becomes more common in hiring, what policies or safeguards should employers put in place to reduce legal and reputational risks?
Employers should start by auditing any AI tool before deployment. They should understand how it works, what data it uses, and whether it has been tested for bias.
It is also important for employers to maintain meaningful human oversight at every decision point, and to document the processes and the rationale behind decisions.
Employers can also ensure that there are adequate safeguards to ensure a fair and impartial review. Revisiting and updating these safeguards regularly are also crucial as AI systems can drift over time.
Ramesh Bharani Nagaratnam, Managing Director, RBN Chambers
Q1 What are the biggest legal risks employers should be aware of when using AI tools in recruitment and hiring?
I believe that the adoption of AI in hiring poses three major risks.
The first is discrimination in the hiring process. AI could filter out candidates based on age, gender, disability, race, or other traits, using details like education history, seniority and gaps in employment. And such discrimination could be difficult to identify until a hard look at the system is made.
The second is data protection. Most AI recruitment tools operate by collecting and analysing personal data such as CVs, online assessments and sometimes video or voice recordings. Such data fall within the personal data protection framework. So it is pertinent for employers or recruiters to obtain the consent of the candidates, clear purpose and strong safeguards in place to ensure that personal data is not violated or misused as otherwise such lapses could result in the infringement of the personal data protection framework and lead to regulatory penalties.
The third is accountability. Though not obliged, when a candidate challenges the outcome of the hiring process, the employer must defend their decision. The employer must be able to provide the basis of their decision to hire or not hire a candidate. Otherwise, word could get around in the labour market that a particular employer engages in “black box hiring”. This could harm the reputation of the employer.
Q2 Could companies face liability if AI systems unfairly filter out candidates or make biased hiring recommendations, even if the technology is provided by a third-party vendor?
In Singapore, recognising the increased use of AI in hiring and the dangers it could pose, under the Tripartite Guidelines on Fair Employment Practices and the upcoming Workplace Fairness legislation, employers are expected to ensure that individuals are hired and treated fairly in workplaces.
So yes, companies can be held liable if the AI (from an external vendor) they use in hiring unfairly excludes candidates or produce biased recommendations. Employers remain responsible for ensuring fair employment practices under Singapore’s Tripartite Guidelines and for complying with the personal data protection framework.
It remains the employers’ responsibility that the AI they use allows for hiring decisions which are fair, transparent and compliant.
Q3 As AI becomes more common in hiring, what policies or safeguards should employers put in place to reduce legal and reputational risks?
In order to reduce the risk that AI could pose in hiring, employers could consider the following:
Assess the risk of AI before it is deployed: For example, identify where AI will be applied, who might be disadvantaged and how critical those decisions would be.
Conduct a due diligence of the AI vendor: Employers could seek from the vendor the disclosure of annual self-assessment, ISO audits and risk assessments, internal assessments, application security and vulnerability testing, penetration testing and encryption of data in transit and at rest.
Put in place operational safeguards: Configure the AI to only rely on job‑related criteria while reserving key decisions regarding a hire to humans. Additionally, the employers must regularly monitor outcomes generated by the AI to ensure that it produces outcomes that align with the employers’ hiring policies.
Transparency with candidates: Employers should disclose to the candidates how AI is deployed in the hiring process and assure that the final decision on hire would be made by the relevant personnel in the organisation not machines.
Training and culture: Equip HR teams to understand AI’s limits and embed its use within existing fairness and data protection frameworks.
Keeping human judgement at the centre of AI recruitment
As AI continues to reshape how organisations attract and assess talent, its role in recruitment is only set to grow. For employers, the appeal of faster screening, improved efficiency, and wider talent reach is clear.
But as the technology becomes more embedded in hiring decisions, so too do the questions around fairness, accountability, and compliance. The insights from HR and legal experts point to a common message: AI can support recruitment, but it cannot replace human judgement or oversight.
Ultimately, the responsibility remains with employers to ensure that AI tools are used in a way that is transparent, well-governed, and aligned with employment and data protection laws. Without the right safeguards, the risks may extend beyond hiring outcomes, potentially affecting trust, reputation, and legal exposure in the long run.
ALSO READ: FWAs in Singapore: The legal must-knows for employers and employees amidst rising costs
share on
