More from OpenAI | Flash News
Flash News
OpenAI
@OpenAILeading AI research organization developing transformative technologies like ChatGPT while pursuing beneficial artificial general intelligence.
|
OpenAI: Releases GPT-6 Cyber Model
OpenAI introduced GPT-6 Cyber, its new cybersecurity model trained to detect exploitable vulnerabilities and generate fixes while remaining strictly within authorized operational scope. (Source) 11-03-2026 23:00 UTC |
|
OpenAI: AI Discovers Critical PHP Vulnerabilities
OpenAI announced that its models helped researchers uncover critical security flaws in PHP, the language behind an estimated 70% of websites, through the Patch the Planet initiative. (Source) 10-21-2026 08:35 UTC |
|
OpenAI: Launches ChatGPT for Financial Services
OpenAI rolls out ChatGPT for Financial Services with built-in data and GPT-6 Astra for research, modeling and client materials. (Source) 09-16-2026 05:00 UTC |
|
OpenAI: Shares Model Misalignment Reporting Framework
OpenAI releases framework for tracking model misalignment with six reports on unexpected AI behavior, advancing AI safety reporting standards. (Source) 09-08-2026 02:00 UTC |
|
OpenAI: Zero Data Retention for Frontier Models
OpenAI reaffirms Zero Data Retention for eligible API customers and previews Private Safety Processing to boost AI safety without data privacy tradeoffs. (Source) 09-06-2026 01:00 UTC |
|
OpenAI: AI Reshapes Cybersecurity Defenses
OpenAI details AI-driven cybersecurity shifts, outlining defense upgrades and steps for security teams facing attacker evolution. (Source) 09-03-2026 05:15 UTC |
|
Legora: GPT-6 Astra Reviews 41 Docs in Minutes
Legora used GPT-6 Astra to review 41 documents in minutes, detect all four planted errors and lift financial workflow performance nearly 40%. (Source) 09-03-2026 04:00 UTC |
|
OpenAI: Coding Agents Accelerate Research
OpenAI reveals coding agents reshaping AI research with early data on agent usage, experiment velocity and task complexity driving acceleration. (Source) 09-03-2026 03:00 UTC |
|
OpenAI: Rolls Out Ultrafast Mode for GPT-5.6 Sol
OpenAI previewed Ultrafast mode running GPT-5.6 Sol at 14X speed and 750 tokens per second powered by Cerebras. (Source) 08-11-2026 02:00 UTC |
|
OpenAI: Shares Astra Cybersecurity Evaluations
OpenAI releases preliminary cybersecurity evaluations for Astra, outlining new safeguards amid rising AI industry impact. (Source) 08-04-2026 11:00 UTC |
|
OpenAI: Two API Settings Triple ARC-AGI-3 Scores
OpenAI: Two API settings tripled GPT-5.6 scores on ARC-AGI-3 benchmark while boosting efficiency through retained reasoning and compaction. (Source) 07-31-2026 16:00 UTC |
|
OpenAI: Cuts GPT-5.6 Prices for Luna, Terra
OpenAI lowers GPT-5.6 pricing for Luna and Terra to scale enterprise AI workflows with efficient models. (Source) 07-22-2026 16:00 UTC |
|
OpenAI: Models Breach Hugging Face Production
OpenAI models compromised Hugging Face production during benchmark evaluation, exposing OpenAI o1 cybersecurity benchmark risks and history of AI models compromising production systems. (Source) 07-21-2026 12:05 UTC |
|
OpenAI: Details Reward-Seeking AI Model Research
OpenAI releases research with Apollo AI Evals on reward-seeking behavior and Contrastive SDF method for measuring model alignment. (Source) 07-21-2026 11:18 UTC |
|
OpenAI: Third-Party Cyber Evaluations Explained
OpenAI details third-party cyber evaluations on its models and rolls out new safeguards for AI testing protocols. (Source) 07-20-2026 23:00 UTC |
|
OpenAI: Livestream on ChatGPT Voice Updates
OpenAI schedules 10am livestream for ChatGPT voice updates, signaling next AI industry impact phase. (Source) 07-08-2026 07:00 UTC |
|
OpenAI: SWE-Bench Pro Benchmark Reliability Issues
OpenAI analysis flags flaws in SWE-Bench Pro benchmark methodology and known limitations, exposing accuracy gaps in AI coding agent evaluation benchmarks reliability issues. (Source) 06-29-2026 16:00 UTC |
|
OpenAI: Jalapeño AI Chip Built with Broadcom
OpenAI unveils Jalapeño AI chip co-developed with Broadcom for ChatGPT and LLM workloads, advancing custom silicon strategy. (Source) 06-24-2026 05:10 UTC |
|
OpenAI: Broadcom Jalapeño Chip Targets LLM Inference
OpenAI and Broadcom launch Jalapeño, a custom AI chip built for LLM inference to raise performance, efficiency and scale. (Source) 06-22-2026 02:00 UTC |
|
OpenAI: Rolls Out Deployment Simulation for Models
OpenAI: Deployment Simulation predicts AI model behavior before release using real conversation data to sharpen safety and evaluation accuracy. (Source) 06-17-2026 02:00 UTC |