The International AI Safety Report, put together by 96 AI global experts, summarises existing literature on the risks and capabilities of advanced AI systems.

Operator and the Computer-Using Agent will be available through the OpenAI API at an undisclosed time.

The White House will ease the way for OpenAI, Oracle, MGX, and SoftBank to build a generative AI computing system.

Explore Grok AI, Elon Musk's chatbot. Is it living up to the hype? Dive into its features, potential, and what sets it apart from competitors.

The tasks feature, now in beta, can provide daily weather reports or send reminders.

The tasks feature, now in beta, can provide daily weather reports or send reminders.

ChatGPT Plus and Pro users now have access to Sora Turbo, intended to be faster and safer than the version shown in February.

Most of the Nova line is available now, with a heavy-duty “reasoning” model coming early in 2025.

AI models within popular tech companies lack transparency in training data, the Australian Senate Select Committee reports.

The AI-generated videos will debut on Instagram next year. Video Gen can create both moving images and sounds. The AI-generated videos will debut on Instagram next year. Video Gen can create both moving images and sounds.

Users on Google Workspace Business, Enterprise, and Frontline plans can now access Google Gemini without switching to a personal email address.

“Copilot is the new UI for AI,” said Microsoft corporate vice president of AI at Work Jared Spataro at the reveal of Microsoft 365 Copilot Phase 2.

OpenAI and Anthropic will give the U.S. government early access to their frontier models for safety evaluation.

Almost no one wants AI to represent people in court – yet. Thomson Reuters Chief Product Officer David Wong explores the ethical use of generative AI in law and what AI can’t do.

Anthropic opened a window into the ‘black box’ where ‘features’ steer a large language model’s output. OpenAI dug into the same concept two weeks later with a deep dive into sparse autoencoders.

How much better is GPT-4 compared to previous models? Learn about cost and capabilities.

Plus, Google reveals plans to unleash Gemini across Workspace to make interpreting long email threads or creating spreadsheets easier.

The production of deepfakes is accelerating at more than 1,500% in Australia, forcing organisations to create and adopt standards like Content Credentials.

Learn how to access Meta’s new AI model Llama 3, which sets itself apart by being open to use under a license agreement.