Broadly, I see following streams of work that AI Governance constitutes:
- Working inside the Government: Some of the ways one can assume roles in Government for AI regulation: legislative branches, domestic regulation, national security, diplomacy, appropriations and budgeting, and policy apparatus assistantships.
- Research on AI policy and strategy
- Corporate AI Governance: Role of non-government institutions - AI industry and non-profits such as GovAI, CAIS, ERO, CLTR, etc. They all have different interests, concerns, vision, and skin in the game. A tech company may prioritize innovation and profitability, meanwhile a civil society organization stresses on human rights concerns in AI.
- On-paper Legal Advocacy
- Off-paper Lobbying: Both advocacy and lobbying can help uncover and bring to light the neglected but promising approaches to governance that have been uncovered in research so far. Lobbyists also act as a bridge between researchers, policymakers, the media, and the public by communicating complicated ideas in an accessible way to many audiences.If done with responsibility, it can influence national narrative on the ethical use and governance of AI systems.
- Technical and Engineering contributions: To bring about progress in AI regulation, we need to design politically accepted methods of AI governance which requires hardware engineers, chip manufacturers, ML engineers for auditing ML models, On-site inspections of models, creating tools or methodologies for identifying and monitoring large-scale data centers, particularly those used for intensive computing tasks, open source developers, information security experts, cryptographers for regulatory agencies, etc. Essentially, a multidisciplinary approach, in supplement to the interdisciplinary approach as often advocated for in AI governance.
- International agreements and multilateral cooperation: to galvanize action toward a stronger and more universal regime. As we realize that China has been a major investor in the AI domain, the hub that fosters many tech MNCs, it is only sensible to build more reliable trust-based relations with such nations given the risks possessed by AI which goes far beyond the nation where it originates.
- Military applications of AI Governance vs civilian applications.
If History teaches us anything: The Internet once used to be a “western-world’s” tool and soon became the thing that drove the world. Revolutions as vast and fast-evolving as these, are soon to detach from the root, thus making their governance, more precisely inclusive governance, critical.
We are, thus, also noticing more south-south cooperation and research collaborations on this front. The trajectory of AI’s future holds transformative potential that could catalyze accelerated socioeconomic progress within the Global South. As scholars committed to this field, it falls upon us, the policymakers, engineers, entrepreneurs, activists to bridge Western AI development with the aspirations of the Global South.