AI Ethics on Trial: The Scarlett Johansson Case and Gender Bias in Technology

For years we have willingly surrendered our identities and personal data to technology companies operating without clear ethical guidelines or public accountability. As technology advances, these companies increasingly assert control over our likenesses, our data and even our sense of agency.

The alleged unauthorised use of Scarlett Johansson's voice by OpenAI's ChatGPT is not just a legal mishap. It is a warning about how AI can exploit our identities and challenge the ethical foundation of our technological progress. Johansson's legal team argued that her personal rights had been violated, and that explicit permission should be required before anyone uses an individual's likeness or voice. The incident raises a broader ethical question about transparency and respect for personal autonomy in AI development.

Authors, artists and other creatives are increasingly clashing with AI companies that often operate under a Silicon Valley ethos of moving fast and breaking things, prioritising innovation over consent and ethical consideration. Johansson's legal action against OpenAI shows how urgently we need robust legal frameworks to protect individuals from such violations. The unauthorised use of her voice breaches both her privacy and her autonomy. This case should prompt policymakers to develop regulations that ensure AI technologies are used responsibly and ethically.

The EU AI Act, which aims to regulate high-risk AI applications and foundation models like ChatGPT, exemplifies the kind of stringent framework needed to address these concerns. The Act mandates transparency, accountability and ethical use of AI, setting a global precedent for AI governance.

If tech companies can manipulate a celebrity's likeness and claim it is synthetic, what prevents them from doing the same to anyone else? This could affect a small business owner whose promotional videos are used without permission, a parent whose family photos are repurposed for advertisements, or a musician whose original compositions are exploited. Ultimately, anyone who values their privacy and identity is at risk.

What each of these incidents reveals is a potentially exploitative relationship between tech companies and creatives. Artists are justifiably concerned that AI companies are selling tools that produce knock-offs of their work without payment or credit. David Holz, the founder of the AI software Midjourney, admitted to using living artists' work still under copyright, without their consent, to train the company's image generation tools. This highlights the broader issue of how AI companies often exploit creative works without proper acknowledgment or compensation. Companies must be held accountable, and clear guidelines should be established to prevent the misuse of AI.

Gender bias in AI

Beyond these ethical violations, the controversy exposes a deeper issue: the gender bias running through AI technologies. Research by UNESCO has found that virtual assistants like Siri and Alexa, which mostly use female voices, reinforce stereotypes of women as subservient and compliant. That is less a neutral design choice than a reflection of the mostly male development teams whose unconscious biases shape these technologies. Studies by the World Economic Forum have shown that women make up only 22 percent of AI professionals worldwide, a disparity that helps perpetuate these biases.

The gender imbalance in the tech industry contributes significantly to the problem. Women constitute a small fraction of AI researchers and developers, leading to technologies that often fail to consider female perspectives adequately. In 2021, the Stanford AI Index reported that women represented only 16 percent of tenure-track faculty focused on AI globally, reflecting a lack of diversity that perpetuates existing biases and results in products that do not serve all users equitably.

Further examples illustrate the broader impact. A review published in PLOS Digital Health highlights that AI-based prediction algorithms often perpetuate bias, leading to disparities in healthcare. Algorithms trained on datasets that do not adequately represent diverse populations can produce less accurate predictions for women and minority groups, which is why gender-sensitive approaches are needed when developing healthcare algorithms. Similarly, an AI recruiting tool used by a global tech company was found to favour male candidates over female ones. Trained on resumes submitted over a decade, it favoured resumes that included words more frequently associated with male applicants, which shows why diverse training data and bias mitigation matter in recruitment systems. Research by the MIT Media Lab found that facial recognition systems have higher error rates for women, particularly women of colour, than for men. These biases can lead to discriminatory practice in security and law enforcement applications, which is why stricter regulation and development practices are essential to ensure fairness.

Reforming AI for an ethical future

The rapid advancement of AI is inevitable, but the direction it takes is not predetermined. To ensure a future where technology serves all of humanity ethically and equitably, we must focus on two critical areas: legal frameworks and education.

Developing comprehensive legal frameworks is crucial to prevent ethical breaches. These frameworks must mandate explicit consent for the use of personal likenesses and enforce accountability. Tech companies should also commit to greater diversity in their development teams, so that AI systems are fair and inclusive. The EU AI Act provides a promising model, aiming to protect fundamental rights and ensure the ethical use of AI across Europe and beyond.

We must also invest in educating the next generation to be ethically minded leaders. Initiatives like Teens in AI inspire young people, women and minorities in particular, and widen diversity in the AI field. Build an inclusive and ethical environment now, and future AI development is far more likely to respect and serve everyone.

These reforms would guide AI development towards a future that honours personal rights, widens diversity and holds companies to ethical standards. Only through real effort in legislation and education can we build technology that serves society rather than exploits it. Bringing a broader range of voices into AI development is about more than fairness. It is also the surest way to build technologies that respect and serve everyone. The Scarlett Johansson case is a reminder that ethics cannot be an afterthought in how we build AI.

This article was first published by Elite Business Magazine.