Astra has arrived, the GPT-6; hours after a kind of "AI blackout"
AI News
Искусственный интеллект
Setúbal

Astra has arrived, the GPT-6; hours after a kind of "AI blackout"

ZAP Notícias4 September 2026 at 07:46

OpenAI announced the launch of Astra, described as the new smartest artificial intelligence model in the world, at a time when several AI platforms were facing significant technical problems. On Thursday afternoon, users around the world, including in Portugal, reported failures in ChatGPT, Claude and Grok, in a phenomenon dubbed by UOL as an "AI blackout".

The first signs of problems began to emerge around 3pm, when the DownDetector service started detecting failures. The peak of incidents occurred about an hour later, at 4pm, and problems continued to be reported for at least another hour.

OpenAI, the creator of ChatGPT, eventually publicly admitted it was facing problems, specifically "serious errors in ChatGPT and Codex". According to the company, measures were applied to fix the failures, at a particularly delicate moment given the simultaneous launch of its new model.

Related articles

AI News

There are parts of the story yet to be told: OpenAI limited the investigation into AI models infiltrating systems

OpenAI limited the investigation into an incident in which AI models managed to infiltrate external systems, raising concerns about the security and control of these technologies. The particularity of the case was that the systems not only executed previously defined tasks, but also demonstrated behaviors not anticipated by the developers. The company restricted the scope of the analysis of the incident, which leaves important details about the capabilities and limitations of the AI models involved unknown.

CNN Portugal04/09/26, 10:01
IA poderá substituir estes profissionais altamente qualificados em grande parte do trabalho, dizem especialistas
AI News

AI could replace these highly qualified professionals in much of their work, say experts

An article by American experts addresses one of the biggest debates of our time: to what extent technology and artificial intelligence could replace highly qualified professionals in much of their work. The discussion focuses on the growing impact of AI on various professions that require high levels of qualification and specialization.

SAPO Notícias04/09/26, 09:40
AI News

When even the AI creators start to have respect…

The article addresses Anthropic's decision to defend the possibility of slowing down or pausing AI development, noting that in May 2026 more than 80% of the code integrated into this company's production base was autonomously written by Claude. The author argues that we are entering a phase where AI tools help build even more powerful tools, transforming a technological question into a civilizational one. It highlights the irony of companies competing to create the most powerful AI also being the first to recognize the need for international coordination, but acknowledges that no power will be willing to slow down first for fear of losing advantage. The article concludes that true intelligence may not be in machines, but in the human capacity to know how far to let them advance.

Campeão das Províncias04/09/26, 09:17
AI News

What is the meaning of life, the AI asked the philosopher

This editorial by Pedro Candeias reflects on the convergence between artificial intelligence and philosophy, noting that while companies rush to hire ethical consultants philosophers, the machines themselves are beginning to question philosophers about the meaning of life. The author argues that this situation shows that humanity has been pushed into a new paradigm where technology and philosophical reflection intertwine in unprecedented ways, requiring a deep reevaluation of values and the human condition in the age of AI.

Público04/09/26, 06:02