Innovating responsibly in an AI-driven era
Training and validating generative AI (GenAI) models is becoming more and more complex, so how can companies developing or operating such systems evaluate and mitigate risks? Funded by the European Innovation Council, the QuantPi project is addressing the problem through its AI Trust Platform, an all-in-one model-agnostic, plug-and-play solution for the risk management of GenAI.
The holistic approach
The platform tackles GenAI model complexity by reducing risks across three critical dimensions: stakeholder alignment, scalable compliance and state-of-the-art technical testing. The dimensions work together to enhance oversight, pinpoint risks, assess risk severity and provide detailed results on which to base effective mitigation strategies. What is important is transparency for all stakeholders. “You can’t address risks you’re unaware of,” explains Lukas Bieringer, head of policy and grants at QuantPi, the German software company coordinating the project. “Our platform’s unique strength lies in its ability to engage all AI life cycle stakeholders effectively in risk mitigation, ensuring informed decision-making and process efficiency without impeding AI transformations.” The QuantPi team recognised the need for a comprehensive approach relatively early on, when market acceptance of an AI system was slow to come due to trust issues related to the ‘black box’ – AI’s internal workings that humans find hard to understand. This is why, unlike other solutions, their platform seamlessly integrates AI testing with robust risk management and compliance needs. “By providing transparent insights and streamlining processes, we empower organisations to navigate and capitalise on AI transformations confidently,” observes Bieringer. The AI Trust Platform has the potential to significantly enhance the reliability and trustworthiness of GenAI systems in a number of fields. One example is autonomous vehicles, where AI systems must make split-second decisions based on sensor data. Another is GenAI systems used for disease diagnoses and treatment planning, where compliance and trust are of the utmost priority. Last but not least are virtual assistants and customer service, where the QuantPi (First automated risk management platform to enable safety, fairness, explainability, and continuous monitoring of generative AI systems) platform can be used to ensure the AI systems and their guardrails protect private information from being leaked, and that virtual assistant and chatbot responses are appropriate, robust and bias-free. If you are interested in having your project featured as a ‘Project of the Month’ in an upcoming issue, please send us an email to editorial@cordis.europa.eu and tell us why!
Keywords
QuantPi, AI, generative AI, GenAI, risk, trust, compliance, risk mitigation, risk management