OpenAI unveils GPT-5.6 family with Sol, Terra, and Luna tiers

After a brief government review delay, OpenAI launched its most advanced model family yet, introducing pricing tiers and a new enterprise workspace tool.

After a brief government review delay, OpenAI launched its most advanced model family yet, introducing pricing tiers and a new enterprise workspace tool.

OpenAI released GPT-5.6 on July 9 with three variants: Sol (flagship), Terra (balanced), and Luna (budget), priced at $5/$30, $2.50/$15, and $1/$6 per million input/output tokens respectively. The company emphasized stronger performance per dollar, improved computer use, and best-in-class cybersecurity capabilities, with Sol achieving 54% better token efficiency on coding tasks than prior versions.

Also debuting was ChatGPT Work, a workplace companion tool combining ChatGPT and Codex for document, spreadsheet, and presentation creation. The release positions OpenAI more directly against competitors Anthropic and Microsoft in the enterprise AI workspace battleground.

SK Hynix completes record $26.5 billion US IPO, largest foreign debut in history

The South Korean memory chipmaker's Nasdaq listing surpassed Alibaba's 2014 record, driven by surging AI server demand for high-bandwidth memory.

The South Korean memory chipmaker's Nasdaq listing surpassed Alibaba's 2014 record, driven by surging AI server demand for high-bandwidth memory.

SK Hynix priced its American Depositary Receipts at $149 on July 9, raising $26.5 billion in the biggest US IPO by a foreign company ever. Shares jumped 14% in their July 10 debut under ticker SKHY. The company remains a key supplier of high-bandwidth memory (HBM) chips for AI servers, with demand reshaping the global semiconductor market.

Microsoft begins rolling out proprietary MAI models in Excel and Outlook

The tech giant is reducing reliance on OpenAI and Anthropic partnerships to gain more control over performance and cost in its productivity suite.

The tech giant is reducing reliance on OpenAI and Anthropic partnerships to gain more control over performance and cost in its productivity suite.

Microsoft's MAI models are now being integrated into Excel and Outlook, marking a strategic shift toward in-house AI capabilities. This move comes as AI commoditization accelerates and companies seek greater control over their AI infrastructure costs and performance characteristics.

Alberta government deploys Claude Code to scan 466 million lines of code for cybersecurity

The Canadian province used Anthropic's model to identify vulnerabilities across government systems in just 20 hours, signaling public-sector adoption of frontier AI.

The Canadian province used Anthropic's model to identify vulnerabilities across government systems in just 20 hours, signaling public-sector adoption of frontier AI.

The Ministry of Technology and Innovation in Alberta used Claude Code with Opus and Sonnet models to audit government code repositories, surfacing security vulnerabilities and building internal tools to harden systems. This represents a shift from pilot projects to large-scale public-sector security work for foundation AI models.