news & insights
back to all news
Author
Etienne Audoin
Partnerships & Innovation Advisor
Partnerships & Innovation Advisor at Arago, Etienne Audoin brings over 20 years’ experience in the HRIS landscape, from PeopleSoft to SAP SuccessFactors, including IBM Watson Talent, where he was one of the first in Europe to champion AI applied to HR for large enterprise clients. With more than seven years at Arago, he leads partnerships with the HR Tech ecosystem and drives product innovation around SAP SuccessFactors and SmartRecruiters. Convinced that agentic AI will fundamentally transform HR practices, he is Arago’s reference expert for AI and Partnerships.
AI in human resources refers to the range of artificial intelligence technologies applied to HR processes, including recruitment, talent management, employee experience, and administrative automation.
As data volumes continue to grow and employee expectations evolve, HR teams are looking to save time on repetitive tasks so they can focus on what truly matters: human relationships and strategy.
This is precisely where AI comes into play, and where Arago has been supporting its clients for several years by combining HR expertise with a strong command of the leading tools on the market.
The first question to ask is not, “Which AI tool should we choose?” but rather, “Which of our processes are the most time-consuming and the least differentiating?” This mapping exercise is often overlooked, yet it shapes everything that follows.
In practical terms, the goal is to identify high-volume, low-strategic-value HR tasks: CV screening, candidate follow-ups, consolidating data from multiple systems, absence management, and reporting. These are the areas where AI can free up time first.
According to McKinsey (2023), generative AI can create approximately 20% of additional value in recruitment and onboarding by improving candidate targeting and significantly reducing the risk of unsuccessful hires.
The processes best suited to an initial wave of automation include:
Deploying AI in HR is not simply a technical decision; it is a strategic one. It involves decisions around data governance, security, GDPR compliance, and, above all, supporting the teams affected by the change.
This is typically where Arago's business consulting expertise comes into play: helping HR leaders structure their AI roadmap, from assessing digital maturity to operational deployment. This approach generally follows four steps:
1. Assessing the current environment: mapping processes, evaluating data quality, and reviewing the existing HR information system.
2. Defining priority use cases: based on potential ROI and technical feasibility.
3. Selecting and integrating the right tools: choosing solutions, configuring them, and connecting them to the existing HR ecosystem.
4. Training teams and managing change: supporting HR teams and implementing relevant performance indicators.
This step-by-step approach helps avoid two common pitfalls: poorly defined AI projects that never get off the ground, and rushed deployments that generate internal resistance.
The solutions available on the market generally fall into three main categories: AI capabilities integrated into existing HR systems, such as SAP SuccessFactors, Workday, or Cornerstone with AI modules; specialised solutions focused on a specific use case (AI-powered ATS, HR chatbots, or talent analytics); and general-purpose AI platforms that can be adapted to HR contexts.
The right choice depends primarily on the digital maturity of the existing HR ecosystem, the size of the organisation, and the complexity of the processes being targeted. A company with 300 employees does not have the same needs or constraints as a group with 5,000 employees.
| Criteria | Traditional ATS | AI-powered ATS |
| CV screening | Predefined keywords | Semantic analysis: understands meaning, not just keywords |
| Candidate-job matching | Manual assessment by the recruiter | Automated predictive matching score |
| Retention prediction | Not available | Yes: based on historical hiring data |
| Risk of bias | Lower, but dependent on the recruiter | Requires careful monitoring and regular audits |
| Productivity gains | Baseline | +35% observed across AI-enabled processes (source: Deloitte, 2023) |
An AI tool can only deliver results if the people using it understand what it does and, equally importantly, what it does not do. Training should therefore go beyond technical adoption and focus on the ability to interpret AI-generated recommendations, challenge them, and make informed decisions.
Companies that invest in developing these skills tend to achieve significantly better results than those that simply deploy a tool and expect adoption to happen naturally.
AI applications in human resources go far beyond recruitment. Here are some of the main use cases adopted by organisations that have embraced AI, as its range of applications continues to expand.
This is often the first use case that comes to mind, and for good reason: AI can deliver immediate and measurable benefits in this area. It can analyse CVs at scale, assess cultural fit beyond technical skills alone, and predict a candidate's likelihood of success in a specific role.
In practice, this means faster pre-screening processes, more relevant shortlists, and recruiters spending less time sorting through applications and more time evaluating candidates.
Onboarding a new employee is a critical phase that can influence both productivity and retention during the first few years. AI can personalise onboarding journeys, automate administrative procedures, and guide new employees through virtual assistants available around the clock.
HR chatbots can answer recurring questions about access to tools, internal processes, or expense reimbursement policies without taking up valuable time from HR teams.
AI can analyse employee potential by combining performance data, declared skills, training history, and behavioural signals. It can suggest relevant internal mobility opportunities before employees start looking elsewhere for opportunities the organisation could have offered them.
This is one of the use cases with the strongest long-term ROI, as the cost of employee turnover and external recruitment can significantly exceed that of a well-managed internal mobility programme.
AI-powered LMS platforms, such as Cornerstone or 360Learning, can analyse skills gaps at both individual and organisational levels and recommend relevant learning paths. Training is no longer a generic catalogue but a personalised experience aligned with business objectives.
According to the WEF's Future of Jobs Report 2025, 39% of existing skills are expected to be transformed or become outdated by 2030 (2025). HR leaders therefore need to anticipate these changes and prepare their teams for jobs that are rapidly evolving.
AI-powered HR platforms can identify early signs of dissatisfaction before they lead to resignations: sentiment analysis of internal surveys, absenteeism patterns, and changes in engagement scores.
These early alerts allow managers and HR teams to intervene at the right time, through a conversation, a training opportunity, or an internal mobility option, rather than simply reacting once an employee has decided to leave.
Document processing, drafting interview summaries, consolidating data from multiple sources, and producing regulatory reports: AI can take over time-consuming and repetitive tasks that consume valuable time without creating significant added value.
McKinsey estimates that AI can reduce the time spent on HR administrative tasks by 20%, potentially freeing up several days each month for a mid-sized HR team—time that can be reinvested in people and higher-value activities.
HR data is among the most sensitive data an organisation handles: health information, performance evaluations, personal circumstances, and trade union membership data. Any processing involving AI must be based on a clearly defined legal basis, employees must be informed in advance, and a Data Protection Impact Assessment (DPIA) may be required where the processing is likely to result in a high risk to individuals' rights and freedoms.
The GDPR does not prohibit AI: it provides a framework for its use. Meeting these requirements is not an obstacle to an AI project; it is a condition for its long-term sustainability.
A DPIA is required under the GDPR for processing operations that are likely to result in a high risk to the rights and freedoms of individuals. It should be carried out before deploying AI systems involving sensitive HR data and documented to demonstrate compliance.
An AI model trained on historical data tends to reproduce the biases present in that data. If past recruitment decisions have favoured certain profiles, AI may perpetuate these patterns, sometimes in ways that are less visible and therefore harder to challenge than a human decision.
Addressing this issue is not optional. It requires regular audits, diverse testing teams, and an internal culture that questions AI recommendations rather than applying them blindly.
AI is a decision-support tool, not a decision-maker. The risk of over-reliance is real: HR teams that simply validate algorithmic outputs can gradually lose their ability to assess candidates and situations independently.
The principle is simple: final decisions regarding recruitment, performance evaluation, or promotion remain a human responsibility. AI provides information; HR professionals make the decisions.
Since August 2024, the European AI Act has been progressively coming into force across organisations using AI in Europe. For HR, the impact is significant: AI systems used for areas such as recruitment, performance evaluation, and workforce management can fall within the Act's high-risk category, which is subject to the strictest requirements.
In practical terms, AI systems used to screen candidates or support employee evaluations may be subject to requirements relating to transparency, documentation, risk management, and formal human oversight. Compliance responsibilities do not rest solely with the software provider: organisations deploying these systems also have obligations under the regulation.
Penalties for certain violations can reach up to €35 million or 7% of global annual turnover. Anticipating compliance requirements when selecting AI tools, rather than addressing them urgently later, is precisely where Arago supports its clients.
Why will AI redefine HR roles by 2030?
The projections point in the same direction: the World Economic Forum and the McKinsey Global Institute both anticipate a profound transformation of the world of work over the coming years.
What this means for HR leaders is that they are no longer simply managers of human resources; they are becoming architects of the transition. Anticipating the impact on jobs, managing reskilling and career transitions, and creating the conditions for confident and sustainable AI adoption are strategic responsibilities, not merely technical ones.
Start with low-risk, high-volume use cases: CV screening, employee FAQ chatbots, and automated follow-ups. These three use cases can deliver measurable benefits within 4 to 8 weeks and allow your teams to gradually build their skills without disrupting the organisation.
Yes, under certain conditions: a clear legal basis, prior information for the individuals concerned, a DPIA for high-risk processing, and guaranteed employees' right of access. The GDPR provides a framework for AI; it does not prohibit it. Arago supports its clients in ensuring compliance for their HR AI projects from the initial scoping phase.
Audit your models using diverse datasets, involve a multidisciplinary team (HR, legal, and data) in their design, and require human validation for any recommendation. Some vendors offer algorithmic fairness certifications, making this an important selection criterion not to overlook.
A traditional ATS filters applications based on fixed criteria (predefined keywords). An AI-powered ATS adds semantic profile analysis, a predictive matching score, and an estimate of the candidate's likelihood of retention, based on the company's historical hiring data.
A pilot project within a defined scope (e.g. recruitment for a specific department) can be launched within 4 to 8 weeks. An organisation-wide deployment in a mid-sized company generally takes 6 to 18 months, depending on the complexity of the existing HRIS and the teams’ level of digital maturity.
Human Resources must not only adopt AI to improve their own function and transformation but also play a role as "guardians of the temple" in supporting and impacting employment in other functions and jobs within the company.
For personalised support in integrating AI into your HR processes, our experts at Arago are at your disposal. We guide our clients step by step to activate the right levers at the right time, depending on the maturity of each company. AI is becoming indispensable, but it is essential to define its scope of action in a personalised way. Do not hesitate to contact us to learn more about using AI in HR.
