Mitigate Machine Bias
Control machine bias to prevent discriminating against your consumers and damaging your organization.
Onsite Workshop
Unmitigated machine biases:
- Are harmful to consumers of your AI-enabled product or service. Such products/services may discriminate against consumers by providing poor quality of service, denying access, or maybe not even recognizing them as customers.
- Are enterprise risk: you are exposing the organization to reputational risk, missed business opportunity, and potentially, litigation and regulatory sanctions.
By identifying and mitigating machine biases:
- You minimize and control their impact, scope, and associated risks.
- You may not be able to fully eradicate machine biases, but you will at least reduce your organization’s exposure and liability while reaping the benefits of AI.
Module 1: Prepare
The Purpose
- Understand your organization’s maturity with respect to data and analytics in order to maximize workshop value.
Key Benefits Achieved
- Workshop content aligned to your organization’s level of maturity and business objectives.
Activities: | Outputs: | |
---|---|---|
1.1 | Execute Data Culture Diagnostic. |
|
1.2 | Review current analytics strategy. |
|
1.3 | Review organization's business and IT strategy. |
|
1.4 | Review other supporting documentation. |
|
1.5 | Confirm participant list for workshop. |
|
Module 2: Understand Machine Biases
The Purpose
- Develop a good understanding of machine biases and how they emerge from human cognitive and societal biases. Learn about the machine learning process and how it relates to machine bias.
- Select an ML/AI project and complete a bias risk assessment.
Key Benefits Achieved
- A solid understanding of algorithmic biases and the need to mitigate them.
- Increased insight into how new technologies such as ML and AI impact organizational risk.
- Customized bias risk assessment template.
- Completed bias risk assessment for selected project.
Activities: | Outputs: | |
---|---|---|
2.1 | Review primer on AI and machine learning (ML). |
|
2.2 | Review primer on human and machine biases. |
|
2.3 | Understand business context and objective for AI in your organization. |
|
2.4 | Discuss selected AI/ML/data science project or use case. |
|
2.5 | Review and modify bias risk assessment. |
|
2.6 | Complete bias risk assessment for selected project. |
|
Module 3: Identify Data Biases
The Purpose
- Learn about data biases: what they are and where they originate.
- Learn how to address or mitigate data biases.
- Identify data biases in selected project.
Key Benefits Achieved
- A solid understanding of data biases and how to mitigate them.
- Customized Datasheets for Data Sets Template.
- Completed datasheet for data sets for selected project.
Activities: | Outputs: | |
---|---|---|
3.1 | Review machine learning process. |
|
3.2 | Review examples of data biases and why and how they happen. |
|
3.3 | Identify possible data biases in selected project. |
|
3.4 | Discuss “Datasheets for Datasets” framework. |
|
3.5 | Modify Datasheets for Data Sets Template for your organization. |
|
3.6 | Complete datasheet for data sets for selected project. |
|
Module 4: Identify Model Biases
The Purpose
- Learn about model biases: what they are and where they originate.
- Learn how to address or mitigate model biases.
- Identify model biases in selected project.
Key Benefits Achieved
- A solid understanding of model biases and how to mitigate them.
- Customized Model Cards for Model Reporting Template.
- Completed model card for selected project.
Activities: | Outputs: | |
---|---|---|
4.1 | Review machine learning process. |
|
4.2 | Review examples of model biases and why and how they happen. |
|
4.3 | Identify potential model biases in selected project. |
|
4.4 | Discuss Model Cards For Model Reporting framework. |
|
4.5 | Modify Model Cards for Model Reporting Template for your organization. |
|
4.6 | Complete model card for selected project. |
|
Module 5: Create Mitigation Plan
The Purpose
- Review mitigation approach and best practices to control machine bias.
- Create mitigation plan to address machine biases in selected project. Align with enterprise risk management (ERM).
Key Benefits Achieved
- A solid understanding of the cultural dimension of algorithmic bias prevention and mitigation and best practices.
- Drafted plan to mitigate machine biases in selected project.
Activities: | Outputs: | |
---|---|---|
5.1 | Review and discuss lessons learned. |
|
5.2 | Create mitigation plan to address machine biases in selected project. |
|
5.3 | Review mitigation approach and best practices to control machine bias. |
|
5.4 | Identify gaps and discuss remediation. |
|