If you don’t think too deeply about it, it sounds like the type of problem-solving gadget that could have useful applications: an app that can change anyone’s accent to a “Western” one (think: non­descript suburban American) in real time. 

That was exactly the intention for Sanas.ai, a Palo Alto–based tech start-up founded by a diverse group of Stanford grads in 2020. The company’s lofty goal, as spelled out in an early press release, is to help “the world understand and be understood, [with] an end goal of unlocking potential through increased understanding and efficacy of communication in digital conversations.” 

This arguably positioned the app as a solution to intentional and not-so-intentional discrimination—by using Sanas, the logic goes, help-desk workers in hubs like India and the Philippines, or immigrants to Western countries, could avoid the mental or emotional anguish that comes with using their real speaking voice. 

It’s just one of the many AI-driven tools that have entered the market in recent years that claim to solve issues related to discrimination at work, or in other social spaces, by erasing people’s identities (that is, accents, skin colour, disabilities and more). FaceApp famously faced criticism for allowing users to lighten their skin tone to make them “hotter,” and more recently added a short-lived selection of new filters (Black, Asian, Indian and Caucasian) so users could “dress up” as different ethnicities. And Sanas.ai is one of many apps promising to “enhance” your natural accent, a euphemism that actually means “erase.”

Many people in the tech community are buying into this idea; Sanas.ai has raised $32 million in Series A funding from investors, including Google and Assurant Ventures, and there are Reddit threads full of people likening these tools to the utopian solution the world has always desperately needed. Apparently, they are the first step to a modern post-racial and post-sexist world!

But that’s not actually what’s happening. Instead, we’re seeing tech companies capitalize on a societal wariness of anyone who looks or sounds different from the perceived norm—something that can be directly tied to our current state of economic uncertainty. Researchers have found that stable, prosperous societies tend to be more inclusive, while those experiencing economic stress can encourage xenophobia, and that’s exactly what we’re seeing right now. An environment of intolerance has been created in many countries by economic downturn and the feeling that things never really recovered after COVID-19. Don’t take my word for it, though. A report in the peer-reviewed journal Nature found ostracism is on the rise and there is a growing apathy to strangers and things viewed as foreign (or even just foreign-ish).

That’s not to say we don’t need a solution. Foreign-born workers in particular have long experienced discrimination during the recruitment process and even once they have been hired, whether it’s because of their credentials, accents, race or religion. In fact, one-fifth of immigrants to Canada have faced discrimination, according to the 2009 General Social Survey conducted by Statistics Canada. This is despite the fact that newcomers are more likely to create new businesses (according to a 2022 study by MIT). North of the 49th parallel, our job-generating entrepreneurs include Aldo Bensadoun (Aldo), Mohamad Fakih (Paramount Fine Foods) and Tobias Lütke (Shopify).

Another meta-analysis found that in a swatch of countries (think: Sweden, Canada, United States, France . . .), minority job applicants received a lower ratio of the percentage of callbacks compared to white native applicants. As someone with a unique Japanese name, this stung a bit, but I can’t say I’m surprised.

Women are also facing similar problems. Company AI hiring systems have been found to prioritize advertising to male workers, and show female ones lower-paying gigs, which has led women to change their names and identities in a bid to trick the algorithm. 

But this is not the first attempt to solve issues of discrimination by encouraging people to become more like the dominant group. Many pre-AI businesses have cashed in on accent neutralization training or skin-lightening products over the decades. And, yet, the world hasn’t magically changed for the better in all the time that these businesses have existed. 

Of course, a new wave of start-ups is here nonetheless, to profit off this, and promising this time it will be different. This time their tools—for a price—will actually make things better. And perhaps we shouldn’t be surprised that tech is leaning into homogeneity on a large scale. The industry has always tilted toward sameness; we watch the same viral videos, use the same phones and download the same apps. But in this case, the promise of erasing your identity doesn’t really work in the long run. 

The core problem for this public frenzy goes far beyond tech; it’s sociological. And unfortunately, an AI tool, no matter how beautifully designed, will never be enough to solve that.