Artificial intelligence hurts women, discriminates in pay and the slow steps of the establishment

She noted that part of the reluctance to legislate stems from the perception that the potential damages are intangible. “Many believe that what exists on the Internet has few consequences in the real world. If a sexual image is not real, how significant can the impact really be? Many would argue that the abuse being carried out through Grok, although grim, does not justify extensive regulation that would restrict the entire sector,” she wrote.

She also noted that during the last few months, very disturbing testimonies of women being secretly photographed by men using “smart glasses” have surfaced. The latest versions, described by Meta as “artificial intelligence glasses,” include built-in tools that generate real-time information about what the wearer is looking at. Feminist organizations have already raised the warning against the possibility of misuse of such technology to identify women and reveal personal information about them.

Meta is actively withdrawing from its supervisory teams. In January 2025, the company announced that it would “dramatically reduce” what it described as “censorship” across its platforms, replace fact-checkers with “community comments” and eliminate “a bunch of restrictions on issues like immigration and gender that are simply disconnected from mainstream discourse.” The company’s own supervisory board criticized the changes, claiming that they were made hastily without transparency as to whether any due diligence was done regarding the potential impact on human rights.

As governments around the world rushed to respond to the wave of millions of offensive images released by Grok, Musk shared a post claiming that Grok’s “popularity and real-world usage” is “skyrocketing around the world.” Abuse seems to be good for business. And this is exactly why regulation is essential.

Only a proactive approach will prevent the collateral damage we are already seeing inflicted on the most vulnerable in our society. But it is also important that regulation not be limited to technology deliberately designed to facilitate harassment. Repeated examples continue to demonstrate how even AI tools with seemingly innocent goals risk perpetuating bias if they are not properly tested before release.

Artificial intelligence discriminates against women – even in salary

In the UK, research by the London School of Economics found that artificial intelligence tools already in use by more than half of local authorities in England underestimated the importance of women in medical conditions, potentially leading to inequalities in care delivery.

Experts have repeatedly warned that artificial intelligence tools used in hiring could actively discriminate against applicants, as well as applicants who wear headscarves and those with “black-sounding names.” However, a study by the University of Melbourne found that 42% of global companies already use predictive artificial intelligence systems in recruitment, with this figure expected to increase significantly over the next few years.

Discrimination – also in health

A study by Global Banking for Women found that artificial intelligence credit rating systems employed by global financial service providers may discriminate against women, risking their exclusion from loans and other financial services, thus exacerbating the already severe global gender credit gap of $17 billion.

Already, as bias and abuse make such tools less safe and enjoyable for women, dramatically fewer women report using AI in their daily lives. The fewer women use AI, the fewer women will feel empowered to contribute to its design, and the more women will be affected by the inevitable job losses it accelerates. The vicious cycle continues.