Confiance.ai launches a new Call for Expression of Interest for innovative startups and SMEs

• Startups and SMEs will have the opportunity to join a group of major industrial and academic players to take part in one of the priority actions of the national artificial intelligence strategy, by advancing their technologies on critical use cases.
• An information meeting will be held online on 7 March, from 10 am to 12 pm for all interested contributors.

The Confiance.ai collective is launching a new Call for Expression of Interest (CEI) for deep-tech startups and innovative SMEs to contribute, through their technologies, to overcoming technological obstacles and to developing a trustworthy environment aimed at industrialising the integration of artificial intelligence into critical and strategic systems (unmanned vehicles, medical devices, robotics, avionics, etc.). The winners will be selected to start work in September 2023.

The Confiance.ai programme is the technological pillar of the Grand Défi public investment programme “Ensuring the security, reliability and certification of systems based on artificial intelligence” launched by the French government as part of the first phase of the National AI Strategy and financed by France 2030. It brings together a multi-sector collective of some fifty partners and benefits from €45 million funding over four years with the aim of providing an environment for the design, validation, qualification, deployment and maintenance of AI-based products and services. The environment will thus contribute to building trust and to the efforts made to move towards the certification of AI-based systems.

A CEI centred on scientific and technological obstacles

The scientific and technological obstacles to be overcome concern data (methods for defining the volume and annotation criteria of data necessary for learning; identification and qualification of biases in data/knowledge), and more generally issues such as access rights and data security, federated learning, and GDPR compliance; the connection between system requirements and the performance specifications to be met by components; the link between the definition of the ODD and the characterisation of the learning, validation and test databases; as well as hybrid methods for improving trust (robustness, explainability, monitoring, etc.) and for NLP or planning use cases.

Increasing the maturity of the winners’ technologies

The selected startups and SMEs will be integrated into one or more of the programme’s projects, enabling them to test and advance their technologies in interaction with the project partners, on use cases proposed by the industrial players. The use cases cover most of the topics of trustworthy AI for critical systems, namely image recognition, speech recognition, natural language processing, time series processing, process supervision and optimisation, embedded target integration, and symbolic and hybrid AI.

The winners will be able to increase the maturity of their solutions and will receive funding from the programme of up to €100,000 per company.

Twelve winners of the first CEI launched in 2021 have joined the Confiance.ai programme consortium. The technologies tested have helped to overcome a certain number of scientific and technological obstacles by enriching learning data, understanding robustness tests and designing explainable algorithms for example.

“Thanks to the Confiance.ai programme, we were able to speed up the go-live of one of our neural network compression methods. In less than a year, we worked on a gas bottle detection use case and halved the inference time of an algorithm developed by Air Liquide. The programme has enabled us to accelerate our developments and has put us in contact with several French industrial players. We were very pleased with our participation and became a full member of the programme at the beginning of the year. I strongly encourage startups developing R&D that can be adapted to the Confiance.ai programme’s issues to respond to the CEI, because they will really save time!”, says Xavier Fischer, CEO of DatakaLab.

“The ambition of making France a “leader in trustworthy AI” is now taking shape through the methods and tools developed by the Confiance.ai programme ecosystem, thanks to the expertise and know-how of major groups, research and academic laboratories, as well as innovative startups and SMEs. The collaboration with startups and SMEs, made possible by the 2021 CEI, has been rewarding for both the programme and the winners. By working closely with industrial players, they have been able to test their technologies and bring them to maturity for the programme’s benefit, thereby meeting certain challenges and overcoming the scientific and technological obstacles to trustworthy artificial intelligence”, adds Juliette Mattioli, Chairwoman of the Confiance.ai programme’s Steering Committee.

Calendar of the Call for Expression of Interest

• February 24th, 2023: launch of the Call;
• March 7th, 2023, from 10 am to 12 pm: online information meeting for all interested contributors
• March 31st, 2023, 6 pm: deadline for submitting responses;
• April and May 2023: interviews;
• June 2023: announcement of selected startups;
• September 2023: start of work;