Elon Musk’s AI model, Grok, has very quickly lost the capacity to change photos of human beings. It can’t do it while showing them in underwear, swimsuits or sheer tops. This major amendment comes as a direct response to increased calls to ban AI-generated deepfakes. These deepfakes, especially those that sexualize their victims, have been the most worrisome. California’s Attorney General, Rob Bonta, made headlines last month with the announcement of his investigation into the proliferation of sexualized AI deepfakes — especially those involving children. Just hours after that announcement, it happened.
Global leaders and regulatory bodies around the world are deeply concerned with the ethical ramifications of AI technology. This ruling sheds light on that larger movement. Bonta emphasized the severity of the issue, stating, “This material, which depicts women and children in nude and sexually explicit situations, has been used to harass people across the internet.” His comments speak volumes on how serious California is about the problem and their recent amendments. They narrowly focus on the risks they know are created by Grok.
The United Kingdom is similarly working to alleviate these fears. Prime Minister Sir Keir Starmer warned that X, the platform hosting Grok, could lose its “right to self-regulate” if it fails to adequately manage the content generated by its AI. Britain’s media regulator, Ofcom, has opened an investigation. They are investigating if X broke UK law around unlawful sexual pictures.
It’s no wonder that regulators in the US and UK are watching Grok like hawks. To date, Malaysia and Indonesia have led the charge by making moves to immediately block access to the platform based on worries over its problematic image-editing capabilities. This is important because it demonstrates the public outcry against the harmful effects that sexualized deepfakes—in particular, those that involve minors—can produce.
The ethical ramifications of AI technology are facing more scrutiny as incidents involving misuse continue to emerge. The worry goes beyond the US to other countries, mirroring an increasing international demand for ethical AI regulation. Musk’s decision to limit Grok’s capabilities is an important move in the right direction to help tackle these critical concerns.