8am☕Coffee

CISOs concentrate a larger share of cybersecurity budgets on software, reaching 40% of total spend, in response to generative AI–powered attacks that execute in milliseconds. The allocation shift prioritizes AI-driven defenses and rapid detection and response capabilities to counter faster, automated threats.

Sakana AI introduces an evolutionary algorithm intended to build stronger AI models without resorting to expensive retraining cycles. The approach aims to raise performance and efficiency while cutting compute costs and time-to-deployment, offering an alternative path to model improvement.

Intuit replaces dependence on basic chatbots with an agentic AI strategy and a repeatable playbook for building task-oriented agents. The framework emphasizes end-to-end outcomes over chat UX and provides patterns others can adapt to deliver more autonomous, reliable workflows.

Eta Compute adopts the ModelCat name and rolls out AI Model Builder, positioning itself squarely around model creation and management. The release signals a focus on simplifying model development and deployment under a unified brand.

Agent Factory details five best practices to strengthen observability for AI agents and raise reliability. The guidance centers on making agent behavior measurable and actionable so teams can monitor, diagnose, and iterate more effectively in production environments.