Artificial Intelligence

This project should therefore have a positive impact on the ability of SMEs to produce and assess quality NLP systems, and should simplify compliance with the AI Act requirements.

With this fellowship, I am addressing the standardisation of AI systems, with particular focus on the standardisation request of the EU Commission in relation to the AI Act. 

My fellowship contributes to the standardisation on trustworthiness and security of AI generative pre-trained models to establish a robust framework that addresses the escalating concerns surrounding the security risks and privacy implications associated with the development, deployment, and utilisation of these advanced AI models. IEEE P7018, CEN CENELEC

With AI’s evolution, a growing need for AI ethicists to address ethical, social, and psychological queries is evident. One gap, however, lies in the absence of standardised competencies for these professionals, causing hesitation among organisations to embrace AI ethics.

The focus of this fellowship is to provide additional practices, guidance and information, for identifying, analysing, evaluating and treating societal concerns and ethical considerations of AI.

These standards will guide SMEs in effectively implementing and benefiting from the advancements in interactive and multimodal AI, ensuring positive integration into their operations and business strategies.

While there are standards describing processes for considering ethical and social concerns in product and service development, such as AI systems, no standard currently addresses the competencies of individuals tasked with implementing these processes in various organisational types.

The draft EU AI Act (including both sets of amendments) lays down significant requirements for logging and record keeping in AI systems.

Harmonised cybersecurity standards for AI systems need to be developed and adopted, as a matter of urgency, for the benefit of the European industry, including SMEs and startups, as well as European users and consumers.

By optimising design, production, deployment and operation of PV, AI can maximise energy generation, leading to increased renewable energy adoption, reducing reliance on fossil fuels.

My work supports AI trustworthiness characteristics such as robustness, human oversight, accuracy, cybersecurity, and transparency (all those are requirements from the AI Act).

In addressing secure AI with the broad interpretation of security to include safety and societal aspects as above TC SAI will engage with EU and other regulatory bodies to ensure that the output supports relevant global, regional, and national requirements.