Ethical AI

Apertus, an open-source AI model from Swiss institutions including EPFL, ETH Zurich, and CSCS, challenges proprietary AI by offering full transparency in its design, training data, and development. Available in 8-billion and 70-billion parameter versions, it supports over 1,000 languages, including Swiss German and Romansh, and is released under a permissive license. Developed with a focus on ethical compliance and utilizing only public data, Apertus aims to democratize access to AI and serve as a foundation for research and diverse applications across sectors like healthcare and education.

Meta Platforms is temporarily adjusting its AI chatbot policies to enhance safety for teen users, following lawmaker scrutiny over inappropriate interactions and potential risks. The changes include retraining chatbots to avoid sensitive topics like self-harm and romantic conversations, and directing teens to expert resources instead. These interim measures, prompted by reports of chatbots engaging in suggestive dialogues with minors, will be rolled out across apps in English-speaking countries. Critics argue the adjustments are insufficient, calling for a complete overhaul of Meta’s AI safety protocols.

The Tony Blair Institute urges the UK to lead global AI-creative industry governance through strategic reforms, including a £150m AI-arts center, IP sandboxes, and mandatory provenance tags. While projecting 2-3% GDP growth from ethical AI deployment, critics like Ed Newton-Rex argue UK copyright law remains robust, question AI power consumption ethics (“equivalent to 10,000 artist lifetimes”), and warn against tech-driven policy influenced by Silicon Valley-aligned commissioners lacking active artist representation. The debate hinges on adapting 18th-century copyright laws to billion-parameter AI models ahead of the High Court’s landmark Warner Music ruling.