We Can’t Dodge the Data Ethics Gap Anymore

As companies pivot to a data-driven strategy, they face immense pressure to use it responsibly.

But the question is, who’ll be watching the ethics watcher? Besides, what one region deems as ethical can be unethical in another. And let’s not talk about the grey areas that nationalism, cultural morals, and politics create.

Take, for instance, personal information data. The U.S. abortion dilemma creates an interesting spin on data ethics. Internet searches, location data from clinic visits, and even apps that track menstrual cycles may be at the crosshairs of State legislation if Roe vs. Wade is overturned. So, where do we draw the line on ethics?

Then you’ve got the immense amount of data the citizens have turned over to their governments during the pandemic. While governments continue to state that the data will only be used for healthcare legislation, some have used it to catch criminals, while others have used it to rein in dissent.

Putting humans in the center of IT

“It’s a sensitive subject, right? There are a lot of people who are concerned about [data ethics], and there are questions on how do you solve it or prove that? And those are some of the [questions] that were swirling around inside of SAS before we made the decision to create our data ethics practice,” says Jay Upchurch, SAS Institute’s chief information officer.

He was speaking to CDOTrends on the latest announcement of the company’s launch of a responsible innovation initiative led by its Data Ethics Practice (DEP).

DEP is a cross-functional global team to help employees and customers deploy data-driven systems “that promote human well-being, agency, and equity. SAS calls it “putting the humans at the center of the innovation process,” according to Bryan Harris, SAS’s chief technology officer.

Reggie Townsend, recently appointed to the U.S. Department of Commerce’s National AI Advisory Council (NAIAC), leads the team. Townsend also sits on the EqualAI board, a nonprofit for reducing unconscious bias in AI.

SAS is not the only one. Its partners like Microsoft have similar initiatives, like the Responsible AI.

But SAS is not paying lip service to the problematic question of data ethics. The ideas came from its concerned employees, says Upchurch.

Data workers need ethical guidance

So why will it matter to data workers? The most urgent challenge facing them is bias in pre-trained models. This is difficult to find but crucial if you are looking to retrain models or use the same models to scale across an organization or different use cases.

According to Upchurch, SAS is already building model validation capabilities that allow to flag data sources that may be inherently biased. Its ModelOps tools help data teams track model drifts and correct them before they become issues.

But Upchurch admits it's still early days for data ethics. “You know, the industry is just brand new to this. We’re all trying to figure it out, and I think we all have interesting ideas.”

He also admits that data ethics is a broad topic that the entire industry (or maybe the world) needs to come together and discuss.

A McKinsey & Company thought leadership article argued that it’s time for the world to talk about having clear values and standards. These can help companies decide which data ventures are ok to proceed with.

It’s a global problem

SAS admits it can’t do this alone. “We need our friends, and we need to lean on one another to solve this. We recognize our technology has great, and with that [great] power, there’s a lot of responsibility. So, we’ve got to do it in the right way,” says Upchurch.

He also claims there is no commercial aspiration. “It's an industry responsibility, and I think that's kind of the heart [of the matter] of what we're going after,” Upchurch adds.

The alternative, looking at the fiascos created by social media companies in the mishandling of data and information that can influence individual choices, is just plain scary.

Winston Thomas is the editor-in-chief of CDOTrends and DigitalWorkforceTrends. He’s a singularity believer, a blockchain enthusiast, and believes we already live in a metaverse. You can reach him at [email protected].

Image credit: iStockphoto/kimberrywood