Growing misuse of AI technology to generate harmful content

Artificial intelligence (AI) systems will increasingly be used to cause harm as they grow more capable. In fact, AI systems are already starting to be used to automate fraudulent activities, violate human rights, create harmful fake images, and identify dangerous toxins.

It has been reported that OpenAI's image generator, DALL-E 3, can be manipulated to create images of children smoking cigarettes. This raises concerns about the potential misuse of AI technology to generate harmful or disturbing content.

Secondly, large language models (LLMs) could be used to increase the speed and scale of text-based cyber attacks such as spear phishing. Frontier language models have the ability to write large amounts of sophisticated text for as little as cents.2 This combination of scale and sophistication could supplant human operators for large-scale spear phishing campaigns.

DALL-E 3 is a powerful AI model that can generate realistic images from text descriptions. While this technology has many positive applications, it can also be exploited to create inappropriate or harmful content. In this case, users have discovered that by carefully crafting their prompts, they can trick DALL-E 3 into generating images of children smoking cigarettes.

This issue highlights the challenges of developing and deploying AI models responsibly. AI developers need to be aware of the potential for misuse and take steps to mitigate these risks. This may involve implementing safeguards, such as filtering out harmful prompts or limiting access to certain types of content.

In addition, users of AI models need to be aware of the ethical implications of their actions. They should avoid using AI to generate content that is harmful, discriminatory, or illegal. It is important to use AI responsibly and consider the potential consequences of their actions.

The incident with DALL-E 3 underscores the need for ongoing dialogue and collaboration between AI developers, policymakers, and the public to ensure that AI is used for good and not for harm.