OpenAI

In a bold display of corporate optimism, Perplexity has handed over the keys to its kingdom, using advanced AI to autonomously write communications, modify software, and monitor production systems with minimal human oversight. * Perplexity is using Astra for end-to-end systems management, drastically reducing the frequency of human check-ins and safety

César de la Fuente's lab has decided to outsource the search for life-saving therapeutics to the same technology currently struggling to count the letter r in strawberry. By deploying OpenAI Codex and ChatGPT for antimicrobial discovery, researchers are scanning both living and extinct genomes in a desperate bid

OpenAI has introduced the ChatGPT Data agent for ChatGPT Work, allowing users to connect company data, uncover insights, and generate interactive dashboards using only natural language prompts. * OpenAI is betting that natural language queries will finally convince corporate executives to trust AI with their actual business metrics. * The feature shifts

In an effort to convince the world it takes guardrails seriously, OpenAI has announced that alignment researcher Paul Christiano is joining the OpenAI Foundation Board and its Safety and Security Committee. * Christiano brings deep technical expertise in AI alignment, safety, and standards to a board historically dominated by commercial ambitions.

In yet another step toward rendering human expertise entirely decorative, an MIT researcher has deployed GPT-5.6 Sol with Codex to autonomously manage quantum computing experiments, crunch the data, and handle qubit calibration without human intervention. * Outsourcing the universe: Quantum mechanics was already confusing enough before we let generative AI

In an effort to bolster independent journalism in Ukraine, OpenAI has partnered with the Association of Independent Regional Publishers of Ukraine (AIRPPU) and WAN-IFRA to launch a dedicated artificial intelligence program aimed at news organizations. * OpenAI provides technology to Ukraine newsrooms under the guise of supporting resilience, ensuring the tech

OpenAI researchers are once again grappling with the philosophical and technical implications of building systems they struggle to fully understand, publishing new reflections on AI alignment challenges. * OpenAI is leaning hard into the 'alien mind' narrative to make standard safety research sound suitably epic. * Calling for international coordination

Financial technology firm Legora recently put OpenAI's latest capabilities to the test, using GPT-6 Astra to conduct a financial statement review across 41 documents in a matter of minutes. According to the reported metrics, the system successfully identified all four intentionally planted errors while boosting workflow performance by

Gaming company Playco has turned to OpenAI's latest technology, using GPT-6 Astra to build three themed game prototypes from a single grey box foundation. According to the company, the process resulted in 50% fewer manual fixes compared to their previous AI workflows, raising fresh questions about the future

OpenAI has announced Daybreak for Frontline Defenders, a one billion dollar initiative designed to expand access to cyber artificial intelligence, specialized training, and technical support for essential services. * OpenAI Daybreak for Frontline Defenders commits one billion dollars to secure critical infrastructure against increasingly sophisticated global cyber attacks. * The funding subsidizes