The world is experiencing an AI rush as organisations push ahead with implementing AI productivity tools, automation systems and other resources in their businesses. The pace is striking: at inDrive, for example, the share of employees actively using AI tools moved from under 20% at the end of 2024 to 80%+ in 2026. The question is: are we truly ready for AI and what it entails in the workplace? Vera Solomatina, SVP of People and Culture at inDrive, points out an often neglected aspect of AI readiness, and the changing role of HR.
Typically, AI transformation is considered solely in terms of the deployment process. Organisations invest heavily in technology, but seem to underinvest in the psychological safety of the people who will actually use it. This is arguably one of the biggest risks in current AI adoption strategies.
In practice, successful AI adoption involves considerable behavioural shifts that depend on whether people feel able to engage with new systems without fear or hesitation. It’s important, therefore, that an organization takes into account the level of internal trust within it when determining AI readiness.
As Yuri Misnik, CTO at inDrive, puts it: the challenge is not any single technical problem, but the transition from a culture of individual AI experiments to a culture of systematic AI operations — one that requires fundamentally changing how teams think about ownership, accountability, competence and what it means to learn on the job.
It’s not just about tool deployment
The gap between installation and integration is where most organisations stumble. Buying an AI platform only creates the illusion of transformation: real readiness is when employees feel safe saying ‘I don’t understand how this works’, and are given the tools and resources to learn.
While much of the corporate response to AI has focused on reskilling at scale, many programmes are missing the point entirely.
Mandatory e-learning and programs designed exclusively for senior or technical staff aren’t usually effective, because people aren’t acquiring the knowledge in a practical, meaningful way. This approach rarely translates into behaviour change or true adoption.
Moreover, the best AI ambassadors are often not the most senior people, and if upskilling only reaches leadership and tech teams, you create an AI-literate elite and a disengaged majority that simply widens the readiness gap.
What works is contextual, embedded learning built into real workflows, not delivered as standalone courses. Within inDrive, we’ve found that stronger outcomes are achieved when learning is integrated into day-to-day tasks and supported through peer-led collaboration.
For example, pairing technical and non-technical employees accelerates the rate of adoption: people are able to learn from colleagues working in different roles and apply those insights directly to their own day-to-day tasks.
At inDrive, an internal network of 50+ AI Champions — employees embedded across every business unit — were trained to create safe spaces for colleagues to ask basic questions, make mistakes, and learn without judgment. Alongside this, we built out more than 30 practical training modules, two formal courses, and a regular mentorship programme.
HR must move upstream in AI decisions
As AI becomes more embedded in how organisations operate, people functions can no longer operate downstream of technology decisions.
HR must move from being a function that responds to AI decisions to being the architect of how AI processes and tools will be implemented, to ensure the well-being and productivity of all teams. This means that the Chief People Officer needs to be at the table when automation decisions are made, not brought in afterwards to manage potential issues.
This shift requires HR teams to develop new capabilities. Data fluency and AI ethics awareness are crucial for the current and future HR role, ensuring HR leaders are able to lead the charge when it comes to AI adoption and the impact on a company.
Look out for old biases
Where AI is used in hiring or to track performance, organisations should also be mindful of bias ingrained in legacy systems that will impact the new. If the data used to train models reflects historical biases, automation doesn’t remove those biases – it scales them. Diversity and inclusion must be embedded in AI design, not added afterwards.
inDrive learned this directly. When we deployed an AI assistant to support our internal Talent Development function, the tool initially delivered around 50% accuracy — not because the model was poor, but because the underlying HR documentation it relied on was fragmented and outdated. Reaching a reliable 90% accuracy threshold required a rigorous, ongoing cleanup of internal records and the creation of approximately 350 curated question-and-answer pairs.
The bottom line is that AI readiness requires planning across these many layers, as well as confidence in using AI tools. This, in turn, means that companies must ensure that trust and an environment of psychological safety are nurtured within the organization’s fabric in the first place.
