https://www.youtube.com/watch?v=M3z0vnVIfpU&t=538s

Despite growing demand for participatory approaches for AI development, there are challenges of ensuring ethical and inclusive stakeholder engagement and preventing “participant-washing” and perpetuating harm to marginalized communities. To support AI development teams and practitioners in working more ethically and responsibly with communities and the public, the Partnership on AI's Global Task Force for Inclusive AI has put forward targeted guidance and resources. These guidelines and resources are designed for those working directly on the design and development of AI-driven technology - particularly within the commercial or for-profit sector. They provide guidance on how to build collaborative relationships with organizations representing, and individuals of, socially marginalized communities.

There are many different ways to solicit feedback from employees, clients, customers, users, or members of the general public. However, they do not all yield the same kind of insights, relationships or benefits. These guidelines and associated guardrails and use cases get into the nuances of building trust-based, mutually beneficial relationships with diverse stakeholders. How stakeholders are meaningfully engaged is just as important as whether they are engaged. Unfortunately, it is possible to build technology that may be socially beneficial, but is also built in an exploitative and harmful manner. It is also possible to establish an ethical working relationship with stakeholders, yet fail to engage them in a manner that leads to the development of a useful AI system.

Given the complexities of both developing AI systems and engaging stakeholders and the broader public, the resource cannot address every possible pain point and scenario. The intention is to provide a broad-but-applicable framework that helps practitioners better understand their unique circumstances when integrating stakeholder input and helps them navigate the common challenges that arise, ensuring that the stakeholder engagement process they design remains authentic and as equity-oriented as possible. For this reason, the guidelines incorporate both guidance and guardrails to help provide an applicable framework with various use cases and examples of the complex scenarios practitioners may be faced with in the process of participatory AI development.

There are substantial limitations to how much benefit can be generated and how much harm can be mitigated. However, the framework provides necessary support for efforts by individuals striving to bring greater alignment between their work and the needs of the communities they wish to serve. This reduces the likelihood of harms and risks those communities may face.

The full draft of the guidelines can be found here: https://bit.ly/PAIGuidelines