Staying on track

Responsible AI Practices

Esther Montoya van Egerschot

Chief AI Governance @ DigiDiplomacy

The comprehensive approach to Responsible AI

 

  • Monitoring: ongoing observation of AI systems

 

A pie chart with three elements: Monitor, Evaluate, and Audit

Responsible AI Practices

Evaluating

 

  • Assessing the effectiveness
  • Assessing ethical performance
  • Measuring outcomes and processes against goals

 

A pie chart with three elements MOnitor, Evaluate and Audit

Responsible AI Practices

Auditing

 

  • Third-party review process
  • Compliance with legal and regulatory requirements
  • Organizational policies
  • Process resembling a financial audit

 

A pie chart with three elements MOnitor, Evaluate and Audit

Responsible AI Practices

Red and blue teaming

Dynamic method to evaluate and improve AI systems continuously

 

  • Red team: challenge the AI
  • Blue team: respond to the red team evaluation

An image of red and blue soccer teams playing against eachother

1 DALL·E
Responsible AI Practices

Continuous evaluation

 

  • Maintain the quality data and systems
  • Check explainability and interpretability
  • Keep it accountable
  • Safeguard rights and stay compliant
  • Scan for any unfairness

 

It's about keeping your AI systems ready and resilient!

A high-tech conference room with professionals analyzing data on large digital displays and computers.

1 DALL·E
Responsible AI Practices

Let's practice!

Responsible AI Practices

Preparing Video For Download...