Actualizado diariamente1509 historias verificadas
Cobertura verificada por IA sobre lanzamientos de productos, cambios en políticas, investigaciones de seguridad y movimientos del sector, explicada en un lenguaje sencillo por un equipo educativo sin ánimo de lucro.
Cada historia enlaza con la evidencia más sólida disponible: fuentes originales cuando están disponibles, y por lo demás reportajes claramente atribuidos.
Qué sucedió, por qué es importante y qué observar: sin impuestos de jerga.
Cuando la señal es escasa, no publicamos nada en lugar de rellenar el feed.
Un flujo creciente de perspectivas verificadas para personas que necesitan comprender la IA sin perseguir exageraciones.
OpenAI says its Astra model can discover and exploit previously unknown vulnerabilities across hardened systems, prompting stronger safeguards and a restricted initial release.
Queensland Law Society’s Proctor reports that Federal Court Justice Michael Lee ordered parties in a Fair Work dispute to investigate whether AI could help organize and analyze evidence, subject to traceability, testing, confidentiality protections and human supervision.
The South China Morning Post reports that Enflame Technology is seeking about US$910 million in Shanghai’s IPO market, testing investor confidence in China’s specialized AI-chip sector.
SecurityBrief Australia reports that Ping Identity launched controls to identify, attribute and govern personal AI agents operating inside enterprise systems.
Researchers introduce CoVA-SFT, a dataset of 51,900 multimodal examples and more than 222,000 reasoning steps, reporting that fine-tuned models more than doubled the performance of interleaved chain-of-thought baselines on its companion benchmark.
Researchers report that an AI agent system organized person-centered temporal clues across dispersed historical texts, achieving higher temporal-localization scores than human-only annotation in six Shiji cases while reducing reported processing time.
A paper accepted to EMNLP 2026 reports that language models represent whether some math and code questions are structurally answerable, yet fail to route that signal into refusal behavior.
OpenAI is urging California Gov. Gavin Newsom to sign Senate Bill 1119, which the company says would require age checks, independent audits, parental controls and protections against harmful AI interactions for users ages 13 through 17.
A paper accepted at EMNLP 2026 reports that inference-time latent-space interventions made large language models give more consistent answers across languages, including gains in English-Arabic and English-Russian factual question answering, without degrading factual accuracy in the tested settings.
A three-seed audit reports that SWIFT generally outperformed confidence-gated layer skipping on accuracy and achieved higher pure inference speed after search overhead was separated. The study also found modest gains but substantial accuracy losses for two trained-routing methods.
A preregistered ablation of a production oral-history transcription tool found that adding full prompt-level context did not detectably improve side-level word error rate on degraded interview audio. The study also reports that pipeline variability exceeded the measured differences, limiting what one transcription…
A new region-aware visual question-answering system reports a modest improvement in detecting unsupported object claims by giving a language model both whole-image and region-level visual evidence.
Un informe útil cada semana
Recibe las noticias verificadas de IA de la semana, datos originales, herramientas útiles, recomendaciones de aprendizaje y nuevos empleos en IA.
¿Contratar a un profesional de IA o lanzar un producto útil de IA? Ponlo delante de personas que han venido aquí a aprender y actuar.
Publicar un trabajo de IA Enviar una herramienta de IA