New AI legislation "creates trust for those who deposit data" in the European UnionPhoto by wildercr on Pexels
Technology
Европейский союз
Искусственный интеллект

New AI legislation "creates trust for those who deposit data" in the European Union

Jornal Económico7 September 2026 at 07:00

Ricardo Teixeira, founder of CompuWorks since 2002 and a specialist with approximately 30 years of career experience started at Microsoft, analyzes in an interview the European Union's Artificial Intelligence Regulation (AI Act), which was approved in August. The regulation creates a legal framework that classifies AI systems into four risk categories: unacceptable risk (prohibited systems such as social scoring), high risk (applications in sensitive areas such as health and education), limited risk (with transparency obligations such as deepfake identification), and minimal risk (common applications). Fines for non-compliance can reach 35 million euros or 7% of annual global turnover.

Teixeira acknowledges that the European Union has a tendency to over-regulate, which can hinder innovation, but argues that the AI Act creates trust and security for those who deposit data in Europe, especially in a context where data sovereignty is increasingly relevant when data is submitted to data centers under US legislation. The specialist cites the recent case of OpenAI, which was creating scoring systems and attacking other systems, as an example of how AI "gained a life of its own."

The interviewee warns that Portugal is very poorly prepared to manage the risks associated with AI, highlighting the low digital maturity of Portuguese companies, particularly among SMEs and family businesses that, according to him, are more concerned with paying bills than with technological security. Teixeira argues that it is necessary to implement rigorous controls, permanent monitoring, supplier policy management, and user training, but acknowledges that it is very difficult to explain these issues to SME managers.

Teixeira predicts that the world as we know it will be "brutally altered" within five years, and that we are already in a phase of "AI against AI," where protection and attack systems use artificial intelligence. Despite considering that the United States is still ahead in innovation, he believes that China is gaining ground quickly and alarmingly, suggesting that leadership may soon change.

Related articles

Technology

Australians may have the option to not use algorithms on social media

The Australian government will present to Parliament a bill that establishes a "digital duty of care," requiring companies like Meta to offer users the option to deactivate the personalization algorithms of their social networks. Communications Minister Anika Wells defended that Australians should have more autonomy over what they see in their feeds, noting that although many appreciate the algorithmic experience, tech giants must respect the choice of those who prefer to disable it.

extra07/09/26, 12:08
Start-up alemã lança primeiro foguetão comercial desde a Europa Continental
Technology

German start-up launches first commercial rocket from Continental Europe

The German startup Isar Spectrum achieved a historic milestone by successfully launching the first commercial rocket from Continental Europe, from a base in northern Norway. The Spectrum rocket carried five European satellites that were placed in Earth orbit, a moment celebrated by the company's CEO and political leaders as the beginning of a "new era" for the European space sector.

SAPO Notícias07/09/26, 11:44
Tecnológica portuguesa lança plataforma que automatiza processos empresariais sem programação
Technology

Portuguese technology company launches platform that automates business processes without programming

The Portuguese technology company MEGATIC, based in Bragança, will launch on September 8th FLOON, a platform that allows companies to automate processes and connect different applications without the need for programming knowledge.

SAPO07/09/26, 11:00
Londres estreia os primeiros robotáxis, mas ainda há um “condutor” a bordo
Technology

London debuts the first robotaxis, but there's still a "driver" on board

Londoners can already use robotaxis through the Uber app since last week, marking the first time this service is available in the British capital through a partnership with a tech company. However, the vehicles still have a safety "driver" on board, indicating that the technology does not yet operate fully autonomously.

SAPO Notícias07/09/26, 11:00