Summary and resources

  • 2 minutes

This module has shown you how to take real responsibility for your bot. It’s not an easy task, but it is worth it. Ethical AI development will ensure that AI becomes a trusted tool to help all of humanity, and bots will go on to assist us in ways that we currently can barely dream of. Remember:

  • Make sure you understand why your bot exists — and that you, and your users, understand the consequences of using it.
  • Make sure your users know they’re working with a bot.
  • Make sure your bot can ask a human for help.
  • Make sure your bot isn’t accidentally evil - and make doubly sure that users can’t make it evil.
  • Make sure your bot works, and that it will keep working.
  • Make sure your bot treats all users fairly.
  • Make sure your bot respects users privacy.
  • Make sure your bot handles data safely.
  • Make sure your bot can be used by the widest range of people possible.
  • You are responsible for your bot’s actions.

Now that you have reviewed this module, you should be able to:

  • Identify potential biases that your bot may contain
  • Ensure your bot respects users privacy
  • Analyze situations where human supervision may be necessary

Use these resources to discover more

Tip

To open a hyperlink, right-click and choose Open in new tab or window. That way you can see the resource and easily return to the module.

Read Lili Cheng's publication Responsible Bots: Ten Guidelines for Developers of Conversational AI.

Consider using the Microsoft Threat Model Tool to aid in developing this for your bot.

Check out Microsoft Inclusive Design toolkit to learn how to design bots that recognize exclusion, learn from diversity, and solve for ability constraints.

The World Wide Web Consortium (W3C) guidelines for web accessibility testing, will help your bot be accessible to the widest possible audience.

You can implement traceability using Azure Application Insights.

Azure AI services like Text Analytics and Content Moderator can help gauge a user’s sentiment, feedback, and identify human handoff points.

Learn more about Microsoft’s approach to AI and take a look at our book, The Future Computed.