The initial compliance deadline for the European Union’s AI Act has recently passed, establishing a risk-based approach to regulate the application of artificial intelligence. This framework prohibits a limited range of “unacceptable risk” use cases, including social scoring that may lead to negative or unfair treatment, as well as the harmful use of “subliminal techniques” for manipulation.
On Tuesday, the European Commission, the EU’s executive arm, released guidance aimed at assisting developers in adhering to this segment of the EU’s AI regulatory framework.
Developers who are eager to ensure compliance with AI applications in the region will find the guidance valuable. Violations of the act’s prohibitions can result in significant penalties, reaching up to 7% of global revenue or €35 million, whichever is higher.
According to a press release from the Commission, “The guidelines aim to ensure a consistent, effective, and uniform implementation of the AI Act throughout the European Union.” However, the Commission also noted that while these guidelines aim to clarify the law, they are not legally binding; enforcement will ultimately rest with regulators and the judiciary.
The Commission further stated, “The guidelines offer legal clarifications and practical examples to assist stakeholders in navigating the requirements of the AI Act,” reiterating its commitment to promoting a safe and ethical AI ecosystem.
Available in draft form, the guidelines can be downloaded here. Their formal endorsement and use are still underway, as the EU is in the process of providing translations in its numerous official languages.
Although the AI Act was enacted last year across the region, its implementation is ongoing, with further compliance deadlines set to begin in the coming months and years.
Enforcement of these regulations will likely take place in a staggered manner — even concerning the prohibited use cases — as EU Member States have until August 2 to appoint the authorities responsible for monitoring compliance with the rule book.
TechCrunch offers a newsletter centered around AI! Sign up here to receive it every Wednesday straight to your inbox.
Compiled by Techarena.au.
Fanpage: TechArena.au
Watch more about AI – Artificial Intelligence
