There is a growing emphasis on the responsible use of AI technology, and how this can be used to address, rather than deepen, bias and diversity, inclusion, and equity issues in society. Highly scalable modern AI systems such as deep neural networks (DNNs) and generative AI are becoming increasingly powerful and capable of generating realistic and convincing content. At the same time, they can be used to target users by recommending highly personalized content. This means that intentionally or unintentionally, they have the potential to be used to create and propagate harmful or misleading content, such as fake news or hate speech.
Responsible AI builds trust and lays the foundation by taking a “human first” approach using technology to help people make better decisions while keeping them firmly accountable through the governance processes and technical steps. There has been growing interest among businesses, policymakers, and researchers in the past few years in issues that focus on making AI technologies fair, ethical, and responsible. Most recently, the Biden-Harris administration has secured voluntary commitments from seven leading AI companies to help move toward the safe, secure, and transparent development of AI technology. This landmark decision suggests that eventually most (if not all) AI-powered technologies will have to showcase their part of being responsible. However, real change and impact can only be achieved when the proposed AI solutions can be scaled.
%Today, 90\% of industry AI initiatives are struggling to move beyond test stages.
Organizations have made tremendous advances in AI, but still fail to integrate their solutions into everyday, real-time decision-making.
This research aims to identify the challenges and opportunities of building responsible AI systems that can be deployed at scale for truthful, fair, and equitable use. More specifically, investigating the answers to the following research questions is of interest:
- How to build AI systems that are representative of the real world and free from bias?
- How to ensure that AI systems present diverse perspectives and offer true facts?
- How to mitigate the risks of AI-powered manipulation and abuse?
- Detecting and mitigating risk and biases in AI at scale?
- How to develop ethical guidelines for the development and use of responsible AI at the organization and national level?