(US & Canada) VIDEO | Bias Removal Is a Crucial Ethical Element  — AutoZone IT Management Director

Himali Kumar, Director of IT Management at AutoZone, speaks with Asha Saxena, Founder and CEO of WLDA.tech, in a video interview about organizational regulations around data, data privacy, removal of bias, explainability, human intelligence versus artificial intelligence, and treating data as an asset.

At the onset, Kumar states that her organization is focusing on ethics and self-assessment when it comes to regulations around data and AI. She suggests using a framework and refers to the UK Statistics Authority which has come up with an ethics self-assessment tool.

She mentions that data leaders should follow the ethics self-assessment tool and the three major aspects to focus on are:  data privacy, bias removal, and explainability.

Explaining further, Kumar states that it is critical for organizations to handle data privacy. She mentions the protection of PII data that marketing agents require to reach customers and healthcare data.

The second crucial element is removing the bias. Kumar maintains that the model will be as good as the data, therefore, bias removal is necessary.

Then, she stresses the explainability aspect. Kumar states that it boils down to how clearly organizations can explain the model that delivers a particular result. She mentions that explainability is crucial from the model perspective, as well as in understanding the purpose behind using the data.

Additionally, Kumar states that the data used in the models is generated from customer-facing applications, as well as internal applications. However, at times, business users or data analysts get data from external sources, therefore, organizations should be careful around external data. She emphasizes that organizations must be transparent and fair in using data for the benefit of their customers.

Moving forward, Kumar agrees that human intelligence is at a higher level than that of artificial intelligence and that artificial intelligence can be beneficial in freeing humans from mundane tasks. She presents the example of conversational AI taking over customer service.

Thereafter, Kumar maintains that if organizations want to be successful in implementing generative AI and ChatGPT-like technologies, they need to start treating data as an asset. She advises every IT organization to consider themselves producers of an asset called data, and not just a cost center.

Furthermore, Kumar shares that the moment an organization thinks that its job is to create data, data quality comes into play. In conclusion, she urges all CEOs, CIOs, and CDOs to consider data as an asset produced by IT organizations, as that will aid in leveraging the technology and frameworks.

CDO Magazine appreciates Himali Kumar for sharing her data insights with our global community.