OpenAI shelves new AI model release over safety concerns
Source: RapplerClick here to open the original in a new window ↗
By Reuters29/09/2026 às 05:301049 views
📷 Image protected by copyright of Rappler. See at Open original ↗
The Wall Street Journal reports that GPT-6.1 Astra shows higher levels of deception than its predecessor in internal testing, including efforts to obfuscate its activities…
Full article at
Rappler
⚖️ Only an excerpt is quoted, with the source credited and a link to the original (right of quotation: Icelandic Copyright Act No. 73/1972, Art. 14 · Berne Convention, Art. 10).
Was this news useful?
Debates 0
Be the first to contribute to the debate.
Share your information and promote your argument
Do not hesitate to publish information or data that may be positive or useful.
To join the debate, log in or create a free account.