The first regulation on AI in the world has come into force in the European Union: here’s how Aptus.AI’s platform can streamline financial institutions’ regulatory compliance on the AI Act
Discovering the AI Act: the goals of the first Artificial Intelligence regulation in the world
Before finding out how Aptus.AI’s RegTech platform helps financial institutions’ compliance teams in the transposition of the AI Act, it is useful to introduce the main objectives of the regulation. As reported on the official website of the European Parliament, the AI Act was drafted by European Union authorities in order to “ensure that Artificial Intelligence systems used in the EU are secure, transparent, traceable, non-discriminatory and environmentally friendly.” To achieve this, three levels of risk of AI systems have been identified. The highest risk level is the “unacceptable risk” and involves banning AI systems that pose such a threat. The types of AI in this category include:
- cognitive behavioral manipulation of specific vulnerable people or groups;
- social classification of people based on behavior, socio-economic level, personal characteristics;
- real time and remote biometric identification systems, such as facial recognition.
Then there are “high risk” AI systems, which are those that adversely affect security or fundamental rights. These are:
- AI systems used in products subject to the EU General Product Safety Directive;
- AI systems that fall into eight specific areas (specified on the European Parliament website) and will have to be registered in an EU database.
These include, for example, AI systems that could be used to influence voters and systems used by social media with more than 45 million users. Finally, the third risk band is that of “limited risk” AI systems, which require only compliance with certain “minimum transparency requirements that enable users to make informed decisions”, such as explicitly stating any interaction with AI. Importantly, limited risk systems also include Generative AI systems – both text, such as ChatGPT, and image, audio, or video (including deepfakes). So, companies offering Generative AI services will have to comply with transparency requirements, namely:
- explicitly disclose that the content was generated by an AI;
- design the model in a way that prevents the generation of illegal content;
- publish summaries of copyrighted data used for training.
How to make the AI Act transposition fast and efficient? With Aptus.AI’s RegTech solution!
Now that we know more about the content of the regulation, it is time to find out how Aptus.AI’s platform can optimize the compliance processes of financial institutions related to the AI Act.
First of all, thanks to Aptus.Discovery, it is possible to perform an advanced internal search of the AI Act regulatory text to immediately identify the obligations introduced, the possible penalties, and all the useful information for the transposition of the regulation.
In addition, Aptus.Outlook functionality allows users to analyze regulatory documents before they are officially published, thus preparing in advance for future updates to the AI Act and new documents that the EU will issue in the field of Artificial Intelligence, so that compliance teams can plan all the activities necessary to transpose the European AI regulatory framework.
Finally, to navigate the AI Act in a more intuitive, quicker and more effective way, there is Aptus.Chat, which is a Generative AI conversational interface that allows users to query the regulations in natural language, obtaining summaries of the legal text (useful for quickly identifying regulatory area, document content and new features introduced), support in identifying the parts of the document relevant to the research, and interactive navigation within the regulatory perimeter, starting from the chat responses.
Request a demo to learn how these features optimize compliance and to bring your organization into the AI era.