On October 25 and 26, at the #InCyber Forum—NORTH AMERICA event, our Confiance IA Chief Executive Officer had the pleasure of moderating the Cybersecurity panel for Trustworthy AI, held at the Palais des congrès de Montréal.
Trustworthy AI is a growing concern for many people today. While there may be some resistance to integrating this still relatively unknown technology, well-designed and implemented AI has already proven to offer incredible and valuable benefits for both individuals and society.
So, how can we design and deliver safe, secure, sustainable, ethical AI? What are the critical links between trustworthy AI and cybersecurity, and how can we rally Quebec’s already well-established cybersecurity and safety expertise around this cause?
At this must-attend event for Quebec’s business and cybersecurity communities, three key themes emerged around cybersecurity and trustworthy AI:
- The areas in which trustworthy AI needs to be implemented and why, in the specific field of learning systems.
- The need to talk about security by design and how it ultimately ties in with the significant end-to-end concepts already seen in the software engineering and data processing lifecycle.
- How to secure prevalent AI tools and applications, such as GPT, by making them more reliable and robust.
The important thing now is to ensure trustworthy AI manufacturing in all its various phases – from the deep engineering of models to the methods and tools used to deploy them on technological platforms. This requires best practices, monitoring metrics and training for AI experts and solution users. Let’s not forget that AI can be used for purposes harmful to systems and humans, such as designing complex attacks, which must be systematically countered by applying a framework and observable metrics. Multi-disciplinary expertise is also essential. Last but not least, it’s also an opportunity to understand how AI can accelerate better cybersecurity practices.
We are delighted to invite you to explore the thoughts and observations shared by our guest experts during this discussion. Thanks to CAE, Polytechnique and CGI for participating and enriching the debate.
Happy viewing!
