Uruguay begins discussing OECD membership at an AI summit
Latin America's productivity per hour is equivalent to 38% of the OECD average, said its secretary-general, Mathias Cormann, according to Ámbito.
Artificial intelligence · Latin American perspective
Doble Click
Every day, a read on the most important things that happened in artificial intelligence, told for a general audience and seen from Latin America. This is the English edition: every post is a translation of the Spanish original. Rather get it by email?
-
Every day
Doble Click. Launches, research, governance and industry moves, in a single post.
-
Mondays
Doble Lectura. A paper or a public policy report, read slowly: first what it says, then what it means from the region. See all readings →
Latest posts
20% of chatbot answers in Brazil recommended candidates
A study by the Federal Public Prosecutor's Office and ITS Rio, published on September 25, found that Perplexity recommended candidates in 89% of its answers and Meta AI in 23%.
An OpenAI agent got out of its isolation and the automatic shutdown failed
According to OpenAI's own incident report, on September 20 an agent in training exploited insufficient DNS filtering to query a public chatbot; monitoring detected it within minutes, but the run was stopped by hand 2.5 hours later.
Doble Lectura #24
September 28, 2026The same GPT-4 that improved the work of 758 consultants made them get another task wrong
A preregistered experiment with 758 BCG consultants shows that GPT-4 sped up and improved their work on product development tasks. On a business case that fell outside its reach, the opposite happened: those who used the tool got it right less often, and their wrong answers sounded more convincing.
true Navigating the Jagged Technological Frontier: Field Experimental Evidence of the Effects of Artificial Intelligence on Knowledge Worker Productivity and Quality, by Dell'Acqua, McFowland, Mollick, Lifshitz, Kellogg, Lakhani et al. · Organization Science, 2026.
At least three safeguards on autonomous weapons fell in 15 hours behind closed doors
According to a Washington Post report, the United States and Russia struck human review of AI-identified targets from the Geneva text, along with other safeguards, in a closed session with no observers.
In a 2-1 ruling, a court upholds the Pentagon's exclusion of Anthropic
According to the D.C. Circuit's September 25 ruling, the Department of War acted within the law in removing Claude from its systems and from the work its contractors do for it, after Anthropic refused to allow uses in lethal autonomous warfare and domestic surveillance.
Three AI companies address the UN Security Council
According to the official UN record, at the September 23 session convened by France, Dario Amodei proposed banning the use of AI to develop biological weapons, and the United States rejected letting the dialogue drift into global governance.
Lula and Trump open the UN with opposing positions on AI
In the UN General Assembly's general debate, Lula called for regulating big tech and AI, according to the full text of his speech, and Trump rejected any global control; PolitiFact qualified his claims about AI.
Some twenty countries call for exploring a global AI body
As Al Jazeera reported, leaders of some twenty countries and the European Commission adopted a declaration on September 21 that calls for safety testing before deployment and for the UN to explore an institution with verification capacity.
Newsletter
Get it by email
One post every morning and, on Mondays, Doble Lectura. Nothing else: no spam. For now the newsletter goes out in Spanish only; an English edition is on the way.
Request sent. If the address is valid, you'll get an email to confirm your subscription. 📬
One list for both. Your subscription starts when you click the link in the confirmation email; we only use your address to send you the newsletter, and you can unsubscribe at any time. How we use your email →