CERTAIN - Centre for European Research in Trusted AI

Some consider the concept of trustworthiness of and trust in AI to be a buzzword or even a category mistake and think it is "only" about reliability and dependability. But we also trust other entities that are inherently beyond formal provability and verifiability, such as institutions, organizations and even processes. The situation is similar with AI.

Kevin Baum, M.A., M.Sc.
Project Manager CERTAIN

People want assurance that the AI systems they use or are affected by will act reliably, safely and ethically. Trust in AI is essential if we are to use this technology in critical applications, whether in medicine, transportation or finance.

Prof. Dr.-Ing. Philipp Slusallek
CERTAIN

BSI supports research and development, as in CERTAIN, to increase both IT security using AI methods and IT security of AI systems. Responsible use of AI systems can only occur if there is independent and verifiable confirmation of the security properties and trustworthiness of the AI systems.

Vera Sikes
Bundesamt für Sicherheit in der Informationstechnik (BSI)

IRT SystemX, as coordinator of Confiance.ai the largest French technological research program in the national AI strategy, welcomes the enthusiastic kickoff of CERTAIN. The cooperation between IRT SystemX and CERTAIN will foster tangible solutions that can be applied in the real world and are ready for deployment in operations.

Lionel Scremin
IRT SystemX

The task of creating trustworthy AI is a collective effort that goes far beyond the boundaries of technology. It requires an interdisciplinary approach where ethics and governance are as important as algorithms and data

Dr. Christian Müller
CERTAIN

CERTAIN stands for "Centre for European Research in Trusted AI" and focuses on an approach that highlights the issue of "trust" in AI systems - an aspect that is often neglected in international research. The aim is to develop new technologies that provide functional and other guarantees for AI systems.

CERTAIN is a consortium, a collaborative initiative involving various partners, legally part of DFKI, focused on researching, developing, deploying, standardizing, and promoting Trusted AI techniques, with the aim of providing guarantees for and certification of AI systems, and it allows for project collaboration with both internal and external partners.

The aim of the CERTAIN consortium is to work across the value chain from basic research to society, focusing on the development, optimization and implementation of Trusted AI techniques to provide guarantees and certifications for AI systems in specific use-cases. In addition to research, the consortium collaborates with industry, standardization bodies and political and societal stakeholders to set certification requirements, define AI trust labels and foster AI literacy.

After establishing itself as a local and regional focal point in Phase 1 and expanding to become a widely visible European lighthouse centre in Phase 2, CERTAIN plans for further growth in the future through the addition of new partners and thematic pillars, as well as the federation of multiple lighthouse centres.

Become part of the CERTAIN initiative by joining the consortium; reach out to us to get involved.

TRUSTED-AI DAY:
KICK-OFF OF THE EUROPEAN CENTER FOR TRUSTED AI

September 19, 2023
DFKI together with DiSerHub and saaris as well as numerous partners from research, industry, authorities, associations, politics and society hosted the first "Trusted-AI Day" in the auditorium of Saarland University. Starting at 12 p.m., the new CERTAIN center of the DFKI was presented to the audience.

things to know …

What is CERTAIN ?

CERTAIN is a consortium, a collaborative initiative involving various partners, legally part of DFKI, focused on researching, developing, deploying, standardizing, and promoting Trusted AI techniques, with the aim of providing guarantees for and certification of AI systems, and it allows for project collaboration with both internal and external partners.

What is the aim of CERTAIN ?

The aim of the CERTAIN consortium is to work across the value chain from basic research to society, focusing on the development, optimization, and implementation of Trusted AI techniques to provide guarantees and certifications for AI systems in specific use-cases. In addition to research, the consortium collaborates with industry, standardization bodies, and political and societal stakeholders to set certification requirements, define AI trust labels, and foster AI literacy.

What comes next in CERTAIN ?

After establishing itself as a local and regional focal point in Phase 1 and expanding to become a widely visible European lighthouse centre in Phase 2, CERTAIN plans for further growth in the future through the addition of new partners and thematic pillars, as well as the federation of multiple lighthouse centres.

How can we join ?

Become part of the CERTAIN initiative by joining the consortium; reach out to us to get involved.

There is a broad consensus that AI models should not only be accurate and efficient, but ought to also have many other properties. The actual compatibility of these properties, however, is not yet well understood, and thus neither is the question of trade-offs and how to resolve them justifiably. CERTAIN's mission is to help provide clarity and certainty.

Kevin Baum, M.A., M.Sc.
Project Manager CERTAIN

We can't just focus on the here and now. We must 'throw the ball far ahead,' in other words, set far-sighted goals that go beyond the current horizon.

Prof. Dr.-Ing. Philipp Slusallek
CERTAIN

We see our role as building bridges between different disciplines to ensure that our AI systems are not only intelligent, but also ethical and socially responsible.

Dr. Christian Müller
CERTAIN

The ability of AI to generate realistic and compelling text, images, and soon videos is impressive, but it poses many dangers. These so-called 'deepfakes' can easily be used to spread disinformation, making it harder for even well-informed people to distinguish truth from – often purposefully spread – fiction.

Prof. Dr.-Ing. Philipp Slusallek
CERTAIN