In June 2020, tech giant IBM announced it would stop offering facial recognition software until further anti-bias testing was conducted.
“IBM firmly opposes and will not condone the uses of any technology, including facial recognition technology offered by other vendors, for mass surveillance, racial profiling, [or] violations of basic human rights and freedoms,” wrote IBM chief executive Arvid Krishna in a letter to Congress.
Shortly after, Amazon and Microsoft made similar commitments, pledging to reassess the use of their facial recognition software by law enforcement due to concerns about bias.
These announcements occurred amidst global anti-racist protests, sparking dialogue about police use of force and racial discrimination within law enforcement and the justice system. This has raised awareness about how bias and discrimination infiltrate daily life, disproportionately affecting marginalized communities.
In response, many companies have committed to combating discrimination by reexamining hiring practices, donating to organizations supporting communities of color, and fostering diversity and inclusivity internally. IBM has gone further by scrutinizing how its technology might lead to harmful — even life-threatening — oversights due to inadequate testing.
Is Technology Biased?
In theory, technology is neutral, incapable of bias or discrimination.
However, in practice, technology often inherits the biases of its creators, whether intentional or not. Without proactive measures to prevent this, its real-world application can alienate users and exacerbate existing issues.
The failures of facial recognition software to accurately identify individuals with non-European features are not isolated. Similar issues have been found in other technologies, even from well-resourced companies, such as Amazon’s AI recruiting tool that was scrapped in 2018 for bias against women.
Ultimately, addressing these biases is crucial for quality. As sophisticated technology reaches a broader audience, your product will likely be used by a more diverse customer base than your development team.
Why risk your product’s ratings, reviews, and your company’s reputation on avoidable issues? Fortunately, there's a solution: test your product with a diverse group of real customers.
Combating Bias with Real-World Product Testing
As we transition into a post-pandemic world, ensuring your product meets diverse needs is more critical than ever. Stress and uncertainty make consumers less tolerant of setbacks. They have more time to research, read reviews, and share detailed feedback. The economic impact of COVID-19 has also heightened the stakes for product launches. Many companies can’t afford a failure.
Ignoring diverse needs creates usability gaps that undermine a product’s success. By diversifying your tester pool to reflect your target market, you can uncover deep-seated issues and create opportunities for product improvement.
Real-world testing provides firsthand data on product performance across a wide range of scenarios. It highlights usability challenges that are hard to replicate in a lab. For example, understanding different accents is a common challenge in voice technology. Real-world testing allows you to identify and prioritize these issues by introducing various accents and assessing how usability changes with environmental factors, like moving from a car to a house with running A/C.
While product makers must actively challenge systemic practices to prevent biases from embedding in new technologies, real-world testing is crucial for avoiding catastrophic oversights now. More importantly, it protects customers from frustrating experiences that erode loyalty and trust.
Get Started with Our Tester Network
For many product testing teams, recruiting diverse testers is a barrier to gaining the insights needed for success. If you struggle to build a tester team that reflects your target audience's diversity, consider using our network of 250,000 pre-profiled users.
