The European Commission released guidelines for AI models categorized with systemic risks to facilitate compliance with the upcoming AI Act effective August 2. Key players like Google and Meta have until next year to align with regulations, which demand thorough evaluation of AI models, risk mitigation, adversarial testing, and cybersecurity measures. General-purpose AI models will also face strict transparency requirements, including the need for comprehensive technical documentation. This regulatory action responds to past criticisms from businesses regarding innovation constraints, reflecting ongoing tensions between regulatory oversight and industry flexibility.
The European Commission issued guidelines for AI models with systemic risks to assist companies in complying with the AI Act active starting August 2, 2024.
Companies like Google and Meta must evaluate their AI models, mitigate risks, conduct adversarial testing, and ensure cybersecurity, facing fines if non-compliant.
Collection
[
|
...
]