Introduction

 

AI has the potential to radically disrupt people’s economic lives in both positive and negative ways. It remains to be determined which of these we’ll see more of. In the best scenario, AI could widely enrich humanity, equitably equipping people with the time, resources, and tools to pursue the goals that matter most to them.

Our current moment serves as a profound opportunity — but only if we put in the effort to work toward positive outcomes.

In medicine and other fields, new innovations are put through rigorous testing to ensure their benefits outweigh probable harms. The AI community, however, has no established practice for assessing the impact of AI systems on inequality or job quality. Without one, it remains difficult to ensure AI deployments are bringing us closer to the economic future we want to live in.

How can we steer AI so its economic benefits are shared by all?

You can help guide AI’s impact on jobs

Partnership on AI’s (PAI) Shared Prosperity Guidelines provide a set of tools to minimize the chance that an AI system worsens shared prosperity-relevant outcomes.

 

There are two parts to these guidelines:

  • High-level Job Impact Assessment Tool for analyzing an AI system’s positive and negative impact on shared prosperity.
  • A collection of Responsible Practices and Suggested Uses for stakeholders able to help minimize the risks and maximize the opportunities to advance shared prosperity with AI.

How to use the Guidelines

This page offers a guided, three-step process for using the Shared Prosperity Guidelines

[STEPS BELOW WILL HAVE GRAPHIC]

STEP 1
Learn about why the Guidelines were created and how to use them.

STEP 2
Apply the High-level Job Impact Assessment to an AI system.

STEP 3

Follow our stakeholder-specific recommendations for Responsible Practices and Suggested Uses.

Get involved

The Partnership on AI seeks to engage all interested stakeholders to test, refine, and drive the adoption of all parts of the Shared Prosperity Guidelines – the Job Impact Assessment tool, Responsible Practices, and Suggested Uses. We also seek to curate a library of learnings, use cases and examples, as well as partner with stakeholders to co-create companion resources to help make the Guidelines easier to use for their community.

We will pursue these goals by means of stakeholder outreach, dedicated workshops, and limited pilot collaborations.

If you’re interested in engaging with us on this work, or want to publicly endorse the Guidelines, you can fill out the Guidelines feedback form, get in touch with us, or sign up to attend the next event hosted by PAI’s AI and Shared Prosperity Initiative.