How AI and voice are perpetuating gender stereotypes – and what you can do about it

Daye Moffitt

Brands must take care not to perpetuate outdated and harmful gender stereotypes as they increasingly utilize humanized voice-activated devices and artificial intelligence (AI) in every part of our lives, Landor’s strategy chief warns.

Speaking at this week’s Ad:Tech conference in Sydney, executive director of strategy for brand consultancy Landor, Daye Moffitt, told attendees the technology revolution is both electrifying and terrifying in terms of how it’s changing how we behave, interact and communicate.

“We know technology is colliding with and changing social norms. What was once reserved for sci-fi is now reality. We are hyper-real, personalized and intelligent,” she said.

However, what’s equally clear is brands are further perpetuating gender stereotypes that have no place in our modern times through these technologies, Moffitt said.

“More and more, we’re tapping into this humanity and realism through these advanced technologies. Just look at the human faces, names and voices being used for robots to make it easier for us to trust them. We’re also letting children play with technology, inviting it into our homes to take control of things for us.

“This technology is playing a key role in social conditioning.”

As examples of how these technologies support gender stereotypes, she noted voice-activated devices in the home, such as Alexa, overwhelmingly use female voices. “Whereas Watson is an alpha male on the executive team,” she pointed out. “Isn’t it disappointing that brands we love are using prescriptive gender types?

“If we continue to do this to technology, we not only reaffirm these stereotypes and sexualization of women, we are perpetuating and amplifying this to the cost of social progress.”

According to Moffitt, there are five things that can help minimize such risk. The first is letting the data decide.

“In some interactions, a female voice makes sense. But the decision should be based on the data, rather than social conditioning,” she said. “For example, Nest was based on data that proved children are more responsive to the maternal voice.”

Moffitt also advised brands to choose to be genderless, and to avoid sitting on the sidelines when it comes to building the skills required to use technology in a culturally acceptable way. She also suggested personifying your brand even as you use AI-powered customer-facing technologies such as chatbots or voice devices.

“And become a mentor – don’t underestimate the power of your own brand,” she said.

Post-presentation, CMO had the opportunity to catch up with Moffitt to further explore how gender bias is still being perpetuated by brands, the pitfalls of data, and her advice on getting your brand ready for the voice era.

One of the challenges with data is that bias is included in a lot of the data sets we used today. Do you have any thoughts on how we try to ensure our data isn’t gender-biased?

Daye Moffitt: One thing I’m always counseling clients on is that so much data is not understood outside one or two data scientists. The value of data is when everyone in the business knows exactly what the story is that it’s telling you. The more people understand the data and are less terrified of it, the more we can determine when the data has been skewed in a way that’s not necessarily right.

We’re still humans and we always say data is only as good as the person evaluating it. It doesn’t always spit out a right or wrong answer, it’s the grey interpretation that’s the magic. The same applies to this – data isn’t necessarily a blanket solution, it should be re-interpreted time and time again.

Are you seeing more democratization and sophisticated use of data across your client base?

Daye Moffitt: I’m a little surprised to be honest, about how backward clients still are on this. It’s a bit different globally when we check into our network. But what we find is a lot of clients come to us with reams and reams of data based on methodologies that aren’t as useful as others – for example, focus groups, online surveys. We have to wade through a lot of it as not much is consolidated into something meaningful. There is a greater opportunity for clients in this market to get smarter with the data they’re pulling – or at lease focusing attention on finding the nugget of gold.

My hypothesis is that as companies and individuals, we’re all expecting too much from the data. We read through, look at the data and expect five answers. If a piece of research or data set identifies one key problem, then I think that’s data being put to good use. Data is also expansive, so unless you can turn it into something you can utilize, then it’s not of great value.

With regards to the wider gender movement, are you finding brands you’re working with are more conscious of addressing stereotypes?

Daye Moffitt: It’s front and center in our minds but at the moment, it’s not on their radar. There is a lot of gender-stereotyping that still exists from a branding perspective. We’ve always referred to boast and planes for example as female, and I can’t help but wonder why. I did ask that question recently when I was on a ferry, and someone said it’s because the ship nurtures us and gets us from A to B. That’s no longer relevant – the women aren’t just simply going to nurture you. Yet we’re still applying those age-old stereotypes on objects brands then put their branding on. If you start unpicking, the world becomes so gender biased.

I think we just need to be conscious of it in every decision we make as brands and individuals, to try not to perpetuate those sorts of stereotypes. I think Nest is a great example here – the company made a conscious decision to use a female voice via data and not because it was just a nice-sounding voice; it had been proven as a more effective piece of technology that way. It’s about the conscious decisions we make and a decision informed not by convention, but because it’s right.

How much interest are you seeing in voice and AI and its role in brand strategy and engagement?

Daye Moffitt: It depends on the category. Finance categories are more heavily focused on those sorts of things from my experience. In times it does come up, our only counsel is to look at what the purpose is for this AI. One thing Landor is dead against is AI for AI’s sake. Unless it has a function for helping the consumer then don’t do it. It has to add value to the customer’s experience.

What advice do you have for brands who are going down that voice-activated path?

Daye Moffitt: Tone of voice is everything for a brand, beyond voice-activated branded experiences. Unless you have a clear, different and interesting tone of voice, you won’t stand out. It’s as important as a visual presence, particularly as the latter starts to erode. That just makes the tone of voice, messages and way you say them become so important.

I’ve always said brand strategy is everything, but it becomes so fundamental as we have to be then very clear on what exactly the purpose of that non-visual brand identity is, the personality. It has to be there in every conversation.