Our Blog
/
Blog

Stakeholder Engagement for Responsible AI: Introducing PAI’s Guidelines for Participatory and Inclusive AI

$hero_image['alt']

While AI technology promises unprecedented capabilities it also presents significant risks and challenges to socially marginalized communities who have historically been excluded from the design and implementation of these transformative systems. While AI systems can make life and work easier, they can uphold and even exacerbate existing inequalities or harm marginalized groups and/or communities. 

At Partnership on AI, we recognize the urgent need for an inclusive approach to the development of AI systems that actively involves people who are most affected by the technology but are often left out of conversations in the development process. To address this disparity we have developed guidelines designed to foster meaningful collaboration between AI practitioners and stakeholders from socially marginalized identities and communities. 

PAI’s Global Task Force for Inclusive AI was launched in 2023 to address the critical challenges of meaningful stakeholder engagement in AI development and deployment. The technology gap, which refers to the gap in who has access to digital tools and who is free from algorithmically driven harm, continues to widen. As such, the need for the involvement of all users, especially those from marginalized communities grows. The Task Force has worked on building upon existing guidance and recommendations to create guidelines that are actionable and ready to adopt for AI developing teams and organizations. 

PAI’s Draft Guidelines for Participatory and Inclusive AI

Over the last year, the Task Force has worked to develop the “Guidelines for Participatory and Inclusive AI.” The resource guides people working directly on the design and development of AI-driven technological solutions through building collaborative working relationships with organizations representing individuals of socially marginalized communities. 

The resource can help practitioners navigate the common challenges that arise in the process so that their stakeholder engagement efforts remain authentic and as equity-oriented as possible. The framework can support individuals’ efforts in aligning their work with the needs of the communities they wish to serve, while reducing the likelihood of harms and risks those communities may face. 

To make these resources as effective as possible, we are seeking feedback from the experts across the PAI community. 

Explore the Guidelines for Participatory and Inclusive AI

Seeking Feedback on the Guidelines

In order to create meaningful resources that have real-world impact we need to ensure that diverse perspectives are incorporated in the development process. This is why we are seeking feedback from:

  • AI developers (and related practitioners that contribute to the development of AI-driven technology) working in commercial, corporate, or other profit-driven settings.
  • Stakeholders from socially marginalized identities and communities or other people who are impacted in some form or another by the development and deployment of technology.
  • Civil society or other not-for-profit entities, aligned with organizational missions to serve and benefit society.
  • Commercial/corporate entities and public-sector government agencies that also adapt and deploy AI technologies at a scale that impacts thousands and millions of lives.

As you review the guidelines, we would especially appreciate your input on the following questions

  1. How can the guidance and guardrails provided for each use case be more descriptive to enable users of this resource to take action and develop more inclusive stakeholder engagement strategies?
  2. Are there any common use cases of stakeholder engagement that you have dealt with that should be included in the guidelines?
  3. Are there additional significant harms to socially marginalized communities or individuals that should be discussed throughout this resource? 
  4. How do you see these guidelines being applied to your work or experience with AI?
  5. What format should the final resource take to ensure you can easily and frequently use it while navigating stakeholder engagement issues?

We welcome feedback from stakeholders across industries and experience and we are excited to foster dialogue with diverse perspectives to develop resources informed by a multistakeholder community. You can submit any and all feedback here! For questions or concerns please reach out at inclusiveai@partnershiponai.org.