After record-breaking negotiations, the EU has managed to square the circle for the world’s first comprehensive set of rules for artificial intelligence.
A number of reports are already circulating about what has been agreed, and following the agreement in principle, further details and technical texts will be worked out in the coming weeks, as it is to be expected that details will still change. It is not expected to be applicable until 2026, and the call for the AI Pact is running in parallel to support companies in planning the measures provided for in the AI Act.
A brief overview of the most important elements of the provisional agreement. Compared to the original Commission proposal, the main new elements of the provisional agreement can be summarized as follows
- Rules for high-impact AI models that could lead to systemic risks in the future and for high-risk AI systems.
- A revised governance system with some enforcement powers at EU level Extension of the prohibited list, but with the possibility of remote biometric identification by law enforcement authorities in public spaces, subject to compliance with safeguards
- Better protection of rights by requiring operators of high-risk AI systems to carry out a fundamental rights impact assessment before an AI system is put into operation.
Further details in the EU AI ACT press release: https://www.consilium.europa.eu/de/press/press-releases/2023/12/09/artificial-intelligence-act-council-and-parliament-strike-a-deal-on-the-first-worldwide-rules-for-ai/
Duty to provide information and public services
Providers of general AI systems such as ChatGPT must provide all necessary information to all downstream economic operators developing an application that falls into the high-risk category so that they can comply with their obligations under the AI Act.
Obligation for public and private entities that provide essential public services, such as hospitals, schools, banks and insurance companies that use high-risk systems, to carry out an impact assessment with regard to fundamental rights.
EU AI Pact
While some provisions of the AI legislation will be applicable shortly after the adoption of the Regulation, others (e.g. some requirements for high-risk AI systems) will only apply after a transitional period (i.e. between the entry into force and the date of applicability).
Against this backdrop, the Commission is initiating the AI Pact, which seeks a voluntary commitment from industry to anticipate the AI Act and start implementing its requirements before the legal deadline.
EU AI PACT Info: https://digital-strategy.ec.europa.eu/en/policies/ai-pact
Start now, introduce and further develop AI
It is therefore advisable for every company that wants to use AI in the corporate context to continuously monitor developments and, in particular, to continuously evaluate possible high-risk applications such as in the HR area.
With our 506 CompanyGPT, we work with our customers to continuously evaluate whether and, if so, in what form an application of the EI AI ACT is necessary, depending on the intended use.