You may remember the #RedrawTheBalance video from eight years ago. It was created by Education and Employers, an independent UK-based charity. In the video, a group of kids is asked to draw a firefighter, a surgeon, and a pilot. Out of 67 drawings, only five show women.

The Global Gender Gap Index is an annual measure evaluating the current status and changes in gender equality across four critical areas: economic participation and opportunity, educational attainment, health and survival, and political empowerment.

Since its launch in 2006, this index has been the most enduring tool for monitoring the progress of various countries in narrowing these gaps over time. The 2024 Global Gender Gap Report predicts that, at the current rate of progress, it will take 134 years for women to reach equal economic, legal, and professional status, as well as equal opportunities in general.

The emergence and rapid evolution of AI-based tools have presented us with a new challenge. AI learns from available data, and we know this data is often biased. For years it has been fed by data reflecting historical gender imbalances, human prejudices, and systemic inequalities.

If the data used to train AI models is skewed or incomplete, the AI’s outputs will inevitably mirror these biases.

When we asked AI to generate images of specific professions, it heavily relied on stereotypical data: A CEO giving a speech? Mostly white men. A kindergarten teacher? Mostly white women in glasses.

Considering the growing popularity of AI tools used for creating content, images, and videos, the concern about widening rather than bridging the gender gap is very real.

Fortunately, where technology can harm, it can — and should— also help to bring positive change.

Give them EqualVoice

The Ringier Group’s initiative EqualVoice was launched in November 2019 by chief financial officer Annabella Bassler. The aim of the initiative is to promote gender equality and increase the visibility of women in the media.

At its core is the EqualVoice factor, which measures the presence of women and men in the media. Today, it is used by 32 media brands in seven countries, reaching 50 million users.

At the EqualSummit on November 12, a new tool called EqualVoice Assistant was introduced to support journalists in the fight for equality. Created as a joint project by diverse teams from Switzerland and Poland, this tool works by scanning content to identify elements that may indicate inequality and reinforce stereotypes. It considers not only who is mentioned but also the context in which individuals are portrayed, the tone used to describe them, and the words associated with them.

For instance, in an article about a president and their spouse, it highlights issues if the spouse’s appearance is the main focus rather than their achievements. The assistant not only points out these areas for improvement but also offers suggestions on how to enhance them.

It is seamlessly integrated into CXP Ring Publishing, so it fits into journalists’ existing workflow. This ensures that, as they develop content, they receive real-time insights on potential gender inequality or stereotypes. With this tool readily accessible, journalists can promptly address and improve content, supporting fair and balanced gender representation in the media.

(It’s worth noting that, despite my best efforts to maintain balanced writing here, the tool revealed biases I hadn’t noticed. It was a clear reminder that, even with the best intentions, we can all benefit from a little help to ensure our content is as inclusive as possible.)

Bartosz Węglarczyk, editor-in-chief of Onet, said, “With the real-time context feature of the EqualVoice Assistant, we will be giving journalists the ability to create content that is not only timely but also consciously inclusive, without compromising on speed.”

You can’t change what you can’t measure

When asked about fighting inequality, Katarzyna Gaweł, head of diversity, equity, and inclusion (DEI) at Ringier Axel Springer Polska, pointed out why it is worth not only taking a series of actions aimed at change, but also checking their effectiveness.

“We believe that our voice is our power,” Gaweł said. “As a media company, we are committed not only to providing content but also to doing so in a responsible and inclusive way that sets the highest standards.

“We recognise the importance of not reproducing stereotypes and countering bias in the content we produce. The language we use should not only reflect reality, but also show different perspectives and shape the consciousness of the millions of Poles we reach with our content. That is why we have been promoting gender equality in our media for many years. In 2020, we started measuring the voice of women in our media with the EqualVoice factor, because you can’t change what you don’t measure.”

In a world where anyone can be a content creator, media bear a tremendous responsibility to ensure that, in the race for speed, they don’t sacrifice important things.

Thay should use any tools they have — their voice and expertise. I hope that, as a result of these efforts, the experiment I mentioned earlier will yield more balanced gender representation in children’s drawings.

Because the content that will be reaching them — directly or through adults — will be much more balanced. You can be what you can see. Let’s help them see broadly, fairly, and without barriers to build a better future.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .