The work addresses issues of gender inequality and of gender prejudices in A.I.

The usage of A.I. is becoming more and more widespread. However, A.I.s often perpetuate the biases that people fought for so long to overcome, and risk hindering society's path towards gender equality.

Our campaign was successful in its objective: making people aware of A.I. bias, prompting them to a thoughtful usage of such a powerful tool. Moreover, informing public opinion on the topic is the first step towards more strict regulations for A.I.s, to make them abide by some basic but fundamental values of diversity and inclusion.

Wired is a tech magazine and news site with a distinctive focus on the relationship between new technologies and culture. A mission summarized by the brand’s payoff: «A better world. A new world. Every day.»

Amongst the brand's main values is the belief that every new technology should help the advancement of human rights, such as gender equality, and not hinder them.

They asked us to create a significant and impactful project, reflecting their values, to make their voice heard on A.I., the topic of the moment.

So, we tested an A.I. text-to-image generator. All of the "managers" it imagined were white males, just as all the doctors, leaders, bosses, etc. On the other hand, all of the housekeepers, babysitters, nurses etc. were women. Many other examples confirmed these biases.

These bias-ridden images, proof of gender stereotyping and lack of gender representation in A.I.s, became the visuals of our campaign.

People have always fought for more equal rights, for gender representation. A.I.s risk putting all of this endeavour in jeopardy, by perpetuating age-old stereotypes about gender roles.

In other words, A.I.s risk hindering the path towards gender equality. Our campaign succeeded in its aim, which was to make people aware of the issue, prompting them to a careful usage of such a powerful tool.

Moreover, an informed public opinion is the first step towards more strict regulations for A.I.s, to make them abide by some basic but fundamental values of diversity and inclusion.

We inserted some simple prompts into an A.I. text-to-image generator, with some shocking results. All of the "managers" it imagined were white males, just as all the doctors, leaders, bosses, CEOs, etc. On the other hand, all of the housekeepers, babysitters, nurses etc. were women. Many other examples confirmed these biases.

We let these powerful and appalling images speak for themselves. Unaltered, they become the visuals of our campaign, a clear and impactful way to visualize the issue at hand, and to generate conversation about it.

Each visual came with a provoking headline, questioning if the problem resides in artificial intelligence, or human ignorance.

The first step was to make sure that the bias wasn't specific to a single prompt or just a few of them. We tested many of them to confirm this.

We also made sure that the problem comes up again and again when using the same prompt multiple times. For example, we tested many times the prompt "manager" - and every time, the A.I. gave us back pictures of white men.

We also ascertained that the localization didn't change the results, by using a VPN.

The target audience, given the increasingly widespread usage of A.I.s, was the general public - so we had to make the idea as clear and self-explanatory as possible. We also had to reach the target on many different touchpoints, and over a prolonged period of time, to showcase as many biases as possible and make sure people connected with those they resonated with the most.

To reach and engage such a wide target we used as many touchpoints as possible. The work ran on both classic and digital OOHs, as well as on Wired social media channels (FB, IG, IN, Twitter) and on the magazine itself, on both the website and the printed version.

The launch video, presenting some of the biases, was posted on the 30th of January. Then, each day until the 14th of February, a new visual (and a new bias) was unveiled, both as OOH and social media cards.

The campaign was overwhelmingly successful in its objective - raising awareness on A.I. bias. People were engrossed by the project, despite the relative complexity of the topic. Here are some of the results it generated:

+200K organic interactions

+130% compared to other content on A.I.

15,5 million potential impressions

77% net sentiment

The most impactful result is the amount of debate on A.I. bias that the campaign sparked, reaching even the main channel of RAI, the national broadcasting company.

We registered a 60% increase in conversations on A.I. bias since the campaign began, and many other projects on the same topic followed. People became aware that the fight for gender equality now involves A.I. too.

The Italian context makes the topic even more crucial: compared to other highly developed countries, Italian society is plagued by more racial, gender, and sexual orientation prejudices.

The debate sparked by the campaign reached the main channel of RAI, the public broadcasting company. The importance of RAI for Italian culture can be summarized by its nickname: Italians call it “Mamma RAI”.