in terms of the flow of data between the United States and the European Union. Enforcement of the GDPR coming next May and alignment of national legislation should be an opportunity for negotiations based on a firmer footing. 2. Consolidating and Raising the Profile of the French AI Ecosystem Europe and France have a high-quality industrial and academic network at their disposal. They are in a position to occupy a central role on the international stage; however, our stakeholders suffer from a genuine lack of visibility. On the other hand, the Asian and North American giants (BATX and GAFAM) guarantee an international reputation for the whole of the ecosystem that they play host to. In the absence of such powerful unifying forces, we must reinforce the connections between stakeholders in our ecosystem with two goals in mind: raising their profile and reinforcing competition in both the domestic and export markets. This lack of visibility is also attributable to the fact that stakeholders in the ecosystem are less able to promote and communicate about their capabilities and their successes. To take an example: whilst the exploits of the major platforms are frequently covered by the mass media, the French company Therapixel’s victory in an international competition in the field of medical imaging was given very little coverage in France. There need to be changes in the rhetoric and in attitudes in order to be able to promote the national ecosystem more widely. Creating a ‘One-Stop Shop’ for Information Relating to AI At least two problems are encountered by potential purchasers of AI solutions: those of formalizing their requirements and identifying the stakeholders who could provide them with a solution. The most common situation goes as follows: a business has a wealth of data history; it would like to make use of this to improve its systems and generate new applications and opportunities. By default, this business tends towards convenience. Understandably, it approaches the international leaders in the field, with their finely-honed rhetoric and their gift for communication, who offers, in exchange for this data, to take on the triple role of providing advice and help with design and development. This is where the problem lies: for the majority of these applications, there are frequently smaller-scale stakeholders who are able to meet their needs with more effective and sometimes less costly solutions. These stakeholders would benefit from being better identified and identifiable, allowing businesses to make a more informed choice. To support future purchasers of AI solutions, it would be advisable to create a ‘one- stop shop’ which could give them advice concerning the nature of their requirements and the stakeholders that it would be appropriate for them to approach. 32