UNLEASH 2025, Paris
Three weeks ago, UNLEASH World 2025 once again took place in Paris. The international meeting point where HR, HR technology providers, and forward-looking perspectives on the future of work come together. As we previously predicted on HRTechArena, this year’s focus had clearly shifted away from experiments and pilots, towards AI in practice: how organisations are truly deploying and embedding artificial intelligence within their processes.
Many organisations, including Microsoft, Novartis and Accenture, shared their experiences of implementing AI at scale. Their stories all carried the same message: success with AI is not only about technology, but equally about data, people, and governance.
From pilots to structural deployment
Where AI in 2023 and 2024 was still viewed largely as a promising experiment, more and more organisations are now moving towards structural deployment and integration across HR and business processes. Several companies presented examples of AI applications in recruitment, learning, workforce planning and HR services. The common thread: AI only creates value when it becomes part of day-to-day work, supported by clear frameworks and committed leadership.
Josh Bersin’s ‘superworker’
Josh Bersin, one of the world’s most influential analysts at the intersection of HR, technology and business strategy, predicted at UNLEASH that 2026 will become the year of the “superworker”. In his view, this is an employee whose capabilities are significantly amplified by AI, not merely supported through automation. According to Bersin, this requires a new organisational model: companies must connect the HR function more firmly to strategic work, job roles and process design. In other words: not just buying tools, but redesigning the work itself.
He also argued that AI agents (and “agentic AI”) will become an integral part of all traditional HR roles, fundamentally reshaping both HR roles and the broader HR-technology landscape.
Trust and responsible use
A recurring theme was Responsible AI. Companies such as Microsoft emphasised that ethics, transparency and human oversight are essential pillars. AI systems can only gain widespread acceptance when employees understand how decisions are made and who remains ultimately accountable. Building that trust appears to be a crucial condition for adoption and long-term success.
The quiet power of data quality
Many organisations highlighted the importance of reliable and consistent data. Poor data quality is, in practice, the biggest barrier to effective AI deployment. As a result, organisations are increasingly investing first in data governance and standardised definitions before rolling out new AI models on a wider scale.
Human capabilities remain central
Another strong theme: upskilling the entire workforce. The adoption of AI stands or falls with employees’ capabilities. Many organisations are now developing large-scale learning programmes that teach employees how to work alongside generative AI, data assistants and automation platforms. The shared message across many UNLEASH presentations: technological innovation only works when everyone understands how to apply it in practice.
Continuous learning and adjustment
Finally, it became clear that AI implementations are never ‘finished’. Models and applications must be continuously monitored, evaluated and updated – not only to improve performance, but also to prevent systems from ageing or developing unintended bias. The most mature organisations therefore treat AI as an ongoing learning journey.
Our conclusion
UNLEASH 2025 marked a clear transition: from ideas and vision to execution. Organisations that view AI as a combination of technology, people and governance are currently making the strongest progress. In the coming years, success will no longer be determined by who has the smartest tools, but by who understands best how to apply them responsibly and sustainably.
