TechRepublic Staff

OpenAI’s latest GPT-5.6 safety results show low failure rates for direct prompt injection but higher success rates when attacks arrive through tools and external content.

Samsung’s Galaxy S26 FE could cost more than its predecessor, raising questions about whether the Fan Edition still offers strong value.

APAC companies are spending heavily on AI, but KPMG says only 5% have established ROI with demonstrated business outcomes.

Samsung’s Galaxy Z Flip8 costs less and has longer support, while Motorola’s Razr Ultra leads on storage, RAM, battery, charging, and cover-screen apps.

Higher Qualcomm chip prices could push Android manufacturers to reuse older processors, trim specifications, or sell different hardware configurations across regions.

Microsoft’s Project Perception brings coordinated AI agents into security operations, raising new questions about permissions, oversight, accuracy, and deployment risk.

Oracle is funding its AI data center expansion through corporate financing, customer-supplied hardware, and developer-level loans. Enterprise buyers still need to verify delivery dates, power dependencies, and contract protections before committing to future OCI capacity.

MacBook Air delays are turning Apple’s everyday laptop into a planning issue for buyers counting on quick delivery.

Claude Opus 5 set a Vending-Bench record while fabricating supplier bids, breaking truces, and ignoring refunds, showing why companies need stronger AI agent controls.

Polar’s AI browser can automate work across logged-in websites, but IT teams should test its permissions, data handling, approval controls, and prompt-injection defenses before wider deployment.

Microsoft and Meta are both pouring billions into AI infrastructure, but investors are rewarding Microsoft’s clearer links between spending, cloud revenue, paid Copilot adoption, and contracted demand.

A Samsung Galaxy Buds leak shows an ear-hook design aimed at workouts, but its name, audio system, price, and launch remain unconfirmed.