Anthropic released Fable, a public but limited version of its cybersecurity AI model Mythos, with strict guardrails blocking many cybersecurity-related requests to prevent misuse. This has frustrated cybersecurity professionals who find the restrictions overly broad, even blocking benign tasks like code reviews or reading blog posts. Fable falls back to a less restricted model when guardrails are triggered. Anthropic is still refining these controls and offers a Cyber Verification Program for approved experts to access fewer limitations. The company aims to balance safety with usability as it evolves its AI security measures.
IBM's business model remains strong despite a 7% decline in mainframe infrastructure revenue, supported by loyal legacy customers and solid free cash flow. Software growth, especially an 11% increase in Red Hat revenue, helps offset mainframe challen...
Energy prices cooled in June, easing inflation pressures despite ongoing tensions between the U.S. and Iran near the Strait of Hormuz. Morgan Stanley reported record quarterly revenue and profit, boosted by a surge in equities trading fueled by the A...
IBM reported earnings and revenue below expectations due to enterprises reducing spending on traditional IT systems to invest in their own AI infrastructure. This shift indicates companies prefer developing on-premises AI solutions with cheaper open-...
US stock futures rose after a softer-than-expected Consumer Price Index (CPI) report reduced the chances of a Federal Reserve rate hike in July. The market now awaits today's Producer Price Index (PPI) data, which could either confirm easing inflatio...
IBM introduced Power Autonomous Operations, an AI agent that autonomously monitors and resolves issues in Power systems up to 15 times faster than manual methods, enhancing system uptime and efficiency. Alongside, IBM launched the Power S1112, a comp...
