NVIDIA's Blackwell Ultra GPUs target enhanced performance for large-scale AI model training and inference.

Google DeepMind initiated its 'AI for the Planet 2026' accelerator, inviting applications from startups applying artificial intelligence to environmental challenges.

A recent experiment demonstrated a fully autonomous AI system executing a ransomware attack, marking a shift in cyber threat capabilities.

Former Federal Reserve Chairman Ben Bernanke has been appointed to Anthropic's Long-Term Benefit Trust, overseeing the AI company's mission and ethics.

Nvidia introduces a financing model, partnering with Coreweave, to grant AI developers compute access with delayed payment.

Google Cloud has made AlphaEvolve, a Gemini-powered AI code-optimisation agent, generally available to customers, enabling automated algorithm discovery as a cloud service.

OpenAI introduces a framework to address national security risks, aiming for responsible AI development and transparency.

OpenAI has consolidated its model architecture around an advanced iteration of GPT-5, significantly expanding its multimodal functions for broader enterprise and consumer applications.

OpenAI launched its GPT-5.6 model following a voluntary submission to a US Department of Commerce review.

Sam Altman's proposal for OpenAI to allocate equity to a U.S. sovereign wealth fund initiates discussion on AI's economic distribution and regulatory engagement.

The US Department of Commerce has removed export limitations on two Anthropic large language models, enabling their global distribution.

Cloudflare mandates AI companies distinguish training crawlers from general web crawlers, enabling publishers to manage data access and potential monetisation.

Samsung Electronics exceeded first-quarter profit estimates, reporting 6.6 trillion won operating profit, fuelled by strong high-bandwidth memory chip sales and contributing to a broader semiconductor rally.

Meta's AI glasses now disable the camera if the recording indicator light is obstructed, enhancing privacy controls.

Illinois implemented new legislation establishing an AI task force and regulatory framework, positioning the state at the forefront of AI governance.

SpaceXAI acquired AI coding startup Cursor for $60 billion and launched Grok 4.5, integrating advanced large language models with developer tooling.

The California Department of Technology has broadened state government access to Anthropic's Claude models, indicating increased public sector integration of advanced AI.

Anthropic researchers have identified a 'global workspace' mechanism within their Claude large language models, suggesting a novel approach to understanding latent reasoning.

Anthropic confidentially submitted IPO paperwork following a fiscal Q3 2026 profit of over $1 billion, indicating strong commercial traction in enterprise AI.

OpenAI provided a restricted preview of its GPT-5.6 Sol model to White House officials, suggesting a growing interface between advanced AI development and government oversight.

A study indicates major AI developers have attenuated voluntary safety measures, potentially increasing risks as capabilities advance.

NVIDIA defers its Kyber gaming GPU architecture launch to 2028, reallocating memory resources to Artificial Intelligence hardware amidst high demand.

Meta platforms now feature an AI image generation tool, prompting user concerns regarding training data utilisation.

The US Cybersecurity and Infrastructure Security Agency now utilises Anthropic's Mythos large language model to conduct extensive code audits for critical national infrastructure.

Microsoft is transitioning to proprietary AI models for internal operations, aiming to mitigate substantial expenditure on external large language models.

Illinois' new Public Act 103-0511 regulates AI in employment video interviews; concurrently, the European Union released a cybersecurity and AI action plan, highlighting diverse regulatory approaches.

Microsoft expands Azure AI portfolio, incorporating Anthropic's Claude 3 models and launching proprietary agentic AI, MAI and Ailon, to foster a multi-model ecosystem.

Microsoft launches 'Microsoft AI,' a programme directing substantial resources towards embedding AI solutions into corporate frameworks, emphasizing wide-ranging enterprise adoption as a core driver for AI value.

The United Nations initiated a global dialogue to establish an artificial intelligence governance framework, prioritising binding international commitments and independent oversight.

Google DeepMind has postponed the Gemini 3.5 Pro release to 17 July, prioritising significant improvements in mathematical reasoning and multimodal integration.