Agency and action: Your role in AI

- Organizations that follow responsible AI practices are able to design and or deploy AI more effectively to achieve high value for customers, employees and shareholders. Therefore, every organization should establish a responsible AI implementation strategy. Such a strategy includes a set of principles like the eight we covered in the first video in this chapter, that guide responsible AI development and deployment practices. For example, an organization may decide to uphold the principle of fairness and nondiscrimination. To implement this principle in practice, the AI implementation strategy may specify that development teams conduct robust testing of an AI system to identify bias, correct these biases before it is deployed and implement continuous monitoring to mitigate harms. Organizations that have a responsible AI implementation strategy will likely be able to respond to new regulations more quickly and efficiently than those that do not. AI regulation is on the horizon. Rather than viewing regulation as a burden, building in compliance mechanisms early can provide a competitive advantage in the long run. We've focused primarily on what it means to be a responsible AI leader at the organizational level. Now, let's consider what you can do as an individual. If your organization has a responsible AI implementation strategy, you should commit to implementing that strategy in all aspects of your work. For example, if your team is considering developing an AI system to streamline a process for its customers, evaluate whether the AI system is in alignment with your organization's responsible AI implementation strategy and especially its responsible AI principles. If it isn't, speak up and identify ways to make it compliant. If it cannot be compliant, you may need to advocate that the AI system not be designed or deployed. If your organization does not yet have a responsible AI implementation strategy, I encourage you to speak to leadership about establishing one. Most large companies such as Google, Microsoft and IBM have a responsible AI implementation strategy that is publicly available. The University of California also has made public its responsible AI implementation strategy that I encourage you to check out. By ensuring your organization has a responsible AI implementation strategy you are helping to make sure that AI systems are designed and deployed in ways that maximize their benefit while minimizing their risks. In the next chapter, I'll share key takeaways and resources for further learning.