With innovation as its main driving force, the digital landscape, especially in the Artificial Intelligence (AI) field, is one of constant change. However exciting, this is a challenge for policy-makers to navigate while trying to capture it and create future-proof legislation. In the case of AI, from data collection and preparation, to model training and testing, to deployment, responsibilities across the value chain are currently not really covering the realities of the digital market. Compliance with existing and overlapping regulation like GDPR and ethical standards must be ensured at all critical touch-points.
In its current form, the AI Act is too simplistic as tries to come up with a “one-size-fits-it-all” solution. Reality is often more complex; it raises more questions. What happens when there are more than one provider? Would a joint providership, similar to the GDPR, be more effective? And more about General Purpose AI: how should it be addressed? What obligations should upstream provider fulfil?