OpenAI Unveils GPT-6 Astra After Its Largest Training Run Yet
OpenAI has started rolling out GPT-6 Astra, a model trained on a record scale, while simultaneously cautioning about its advanced cyber capabilities.
Latest artificial intelligence news, tools, models and updates.
OpenAI has started rolling out GPT-6 Astra, a model trained on a record scale, while simultaneously cautioning about its advanced cyber capabilities.
Family-focused AI assistant Ollie is positioning itself as a privacy-conscious alternative to mainstream assistants, pledging not to train AI models on user data or share it with third parties.
Major AI platforms including ChatGPT, Claude, and Grok are experiencing widespread outages, disrupting users globally. Here's what we know so far.
OpenAI's Astra is the first model to meet the Critical cybersecurity capability threshold under the Preparedness Framework, signaling a new era of AI safety and frontier safeguards.
Anthropic rolls out AI agent blueprints for retailers amid holiday season, while its IPO prospects and security transparency demands draw attention.
A new METR report reveals how a Hugging Face breach and OpenAI's agent swarms highlight systemic security weaknesses in AI development, urging stronger supply chain defenses.
OpenAI's upcoming Astra model will employ a novel technique called 'recurrent depth,' which departs from the sequential reasoning typical of most AI models. This unconventional approach has sparked alarm among AI safety experts.
OpenAI's ChatGPT can now connect to trusted healthcare data, enabling clinicians to securely access patient context and medical research within the AI assistant.
Basis, Clay, and Exa Labs demonstrate how AI agents enhance onboarding, account management, and developer integrations, offering a blueprint for enterprise leaders.
Multiverse Computing unveils Quasar 438B, a 438-billion-parameter open-source model trained in Europe, challenging US and Chinese AI leadership.