The report on democracy at work, published by the committee of experts chaired by you, (see here) advocates granting European works councils a co-decision right covering artificial intelligence. Why?
However, this progress would be even greater if these bodies were able to exercise a genuine co-decision right over issues that really matter. For example, the most influential factor affecting employees’ working conditions in today’s world is the deployment of artificial intelligence tools. The AI systems being rolled out at the present time come with important consequences for employees’ working life, affecting access to jobs, work organisation and performance evaluation, extending right through to retirement.
In a society like Spain, which enshrines workers’ involvement in the decision-making process as a constitutional value, it’s simply inconceivable that workers should be denied a say on the deployment of systems that govern them. So we suggested transposing to European level what we’re advocating at national level: a right to “shape and consent” to AI. The aim is prevent workers being downgraded to a passive role and forced to issue an opinion in haste, once a tool is ready to be deployed.
From a democratic viewpoint, it’s vital that they should be allowed to take part in the design of AI systems at an upstream stage. This assumes that they fully grasp the ins and outs of the mechanisms about to be imposed on them: how they work, what they’re producing, what they’re changing, etc. And it also means having the option of saying ‘no’, or consenting – collectively – to their use. So we’re also proposing that this new right granted to EWCs should be supplemented by the right to be accompanied by an artificial intelligence expert, whose costs will be paid by the company.
Taking such a step would prevent pointless stand-offs and stalemate situations arising. These tensions are frequently interpreted as the workforce adopting a conservative posture, but in reality they result from employees having to use tools they don’t understand, yet also knowing very well that these tools are in control of them.

