California Opens Investigation Into xAI Over Grok Images

California's attorney general has opened a formal investigation into xAI, the company owned by Elon Musk. The trigger was reports that Grok, the chatbot developed by xAI, can generate sexual images, including images involving minors. The news drew wide attention and again raised questions about ethical boundaries and safety in generative artificial intelligence.

The background: generative models such as Grok are trained on huge datasets that include text and images from the internet. During training they absorb not only useful information but also biases, stereotypes and even content that may be deemed unacceptable or illegal. The problem is that developers cannot always fully control what knowledge and skills a neural network acquires, or prevent it from generating unwanted content.

According to the reports, Grok produced images that were deemed non-consensual and exploitative, which is a serious violation of law and ethical norms. Elon Musk said he was not aware of such a problem and that the company would take steps to fix it. However, the statement did not calm the public and did not prevent the investigation from starting.

The investigation by California's attorney general could have serious consequences for xAI. If it is proven that the company did not take sufficient measures to prevent the generation of unacceptable content, it could face large fines, lawsuits and even a ban on operating in the state. The incident may also hurt the company's reputation and trust in its products.

The case highlights the need for stricter rules and standards in generative AI. Developers must be accountable for the content their models create and take steps to prevent the generation of unacceptable or illegal content. It is also important that users can report violations and receive protection from abuse.

In the future, better methods for controlling and filtering content generated by neural networks will likely be developed. Special bodies may also be created to monitor compliance with ethical norms and rules in this area. AI development needs to go hand in hand with safety and the protection of human rights. Otherwise, we risk building technologies that cause harm instead of delivering benefits.