Toms Hardware reports that Nvidia's seven-year-old Shield TV Pro has received a 50% price increase amid an AI-driven memory shortage. The new $299.99 price makes the long-running streaming device substantially more expensive than at launch, illustrating how component supply pressure can reach older consumer hardware.
BleepingComputer reports that Anthropic has begun asking Claude users to voluntarily share voice conversations to help train and improve its AI models. For teams using voice assistants at work, the request makes consent and the handling of sensitive spoken information relevant considerations.
The Hacker News reports that China-linked espionage group TA419 has targeted AI experts at US think tanks, universities and legal organizations with credential phishing. The campaigns impersonate economists, AI policymakers and an Anthropic employee, showing how credible professional identities can be used to target specialist communities.
InfoQ reports that CVE-2026-85706, a critical path-traversal vulnerability affecting self-managed GitLab CE and EE, is being actively exploited. The reported flaw could let an unauthenticated remote attacker read arbitrary files from the GitLab instance, making exposure assessment a priority for teams operating their own installations.
Developers working at AWS have open-sourced Pizza Bot, a self-hosted application that lets AI agents run background tasks and return results through an inbox interface. According to InfoQ, it supports scheduled and webhook-triggered work, delegation to specialized workers and pauses for human approval.