In February, representatives of the 27 member countries of the European Union approved the final text of the first law on Artificial Intelligence (AI-ACT).
The European Union will establish rules for the development and deployment of AI through law, with the aim of preventing or minimising harm. AI-ACT also aims to harmonise the market for AI and promote its adoption in the EU.
The AI law still needs to undergo formal approval from the European Parliament, but its implementation will be key to ensuring that AI regulations do not overburden companies.
Stay up to date with the main points of this legislation
The European Union aims to regulate AI as part of its digital strategy, creating better conditions for the development and use of this telegram data technology. So, check out some of the points of the new AI legislation:
Who will be affected by the legislation: Firstly, the law will impact people in the European Union and society as a whole, as AI systems permeate our daily lives. And secondly, companies and authorities that provide or use AI systems in the region, regardless of where they are based.
Risk-based approach
AI-ACT follows a risk-based approach, categorizing systems into high-risk, subject to specific rules, and low-risk systems, with fewer obligations.
Transparency and compliance for AI models: The agreement requires basic AI models, such as ChatGPT, to comply with transparency the impact of anchor text on long-tail keyword rankings before being made available on the market. This includes producing technical documentation, complying with EU copyright laws, and disclosing information about the content used in training.
Protective measures applicable to generative AI systems:
Developers of generative AI systems and their associated models must follow transparency guidelines, including technical documentation and compliance with EU copyright laws. They must also disclose details about the data used in training.
>The new law also establishes stricter measures for high-risk generative systems, including risk assessment and mitigation, testing, major incident reporting, and cybersecurity assurance.
Risk assessment and management: High systemic risk AI models require assessments, analyses, risk mitigation and adversity testing. Therefore, when a serious incident is identified, it must be reported to the European Commission in order to ensure cybersecurity.
Prohibitions and consumer rights: The agreement prohibits manipulation of human behavior, indiscriminate deletion of facial images and creation of social score rankings.
Possible impacts of AI-ACT in Brazil
Although AI-ACT is European legislation. Its eu phone number scope may affect companies operating in the EU or offering products and services to European consumers.
Therefore, Brazilian companies that develop or use Artificial Intelligence in the European market will need to ensure the compliance of their systems. This may require important revisions to operations. In addition, it will also serve as an important model for Brazil, which is beginning discussions involving AI legislation.
In summary, the EU’s approval of AI-ACT marks a significant step forward in the regulation of Artificial Intelligence. It aims to protect citizens, promote transparency and harmonize the market.
The new law establishes risk-based criteria and specific measures for different types of AI systems. It thus sets a standard for the responsible development and use of this technology.