Paedophiles using open source AI to create child sexual abuse content, says watchdog
The widespread availability of open-source artificial intelligence allows offenders to generate child sexual abuse material, creating a severe safety challenge. Unlike proprietary tools, these models can be downloaded, modified, and run locally, enabling users to fine-tune algorithms using illegal images. This accessibility means that once the technology is released, it cannot be recalled or effectively contained, leading to an irreversible proliferation of tools that facilitate the creation of photorealistic abuse content. This development significantly complicates efforts to protect victims and manage the volume of harmful material. The technical community within the offender space actively shares models and tips to enhance realism, often starting with publicly available foundational images. Consequently, law enforcement and child safety organizations face an overwhelming burden as AI-generated content makes it harder to identify real-life victims and coordinate removals, exacerbating an already critical public health epidemic. This situation is highly relevant to open data and open source principles, as it highlights the inherent trade-off between technological democratization and security. While open-source software fosters innovation and prevents corporate monopolies, it also removes barriers for malicious actors, posing unique regulatory challenges. It underscores the difficulty policymakers face in balancing the benefits of transparent, accessible technology with the urgent need to prevent its misuse in creating and distributing illegal content.
Source: theguardian.comPublished on 2023-09-13