NIST’s AI Guardrail Proof Makes Prompt Injection a Continuous Security Problem
NIST says a fixed set of AI guardrails cannot be universally robust against adaptive adversarial prompts. For teams deploying AI agents, the practical answer is continuous red-teaming, guardrail updates, access limits, and recovery planning.
NEURA’s $1.4B Raise Turns Humanoid Robots Into an AI Infrastructure Race
NEURA Robotics raised up to $1.4 billion from backers including NVIDIA, Amazon, Qualcomm, Bosch, Schaeffler, Tether, and the European Investment Bank. The funding makes humanoid robots look less like a gadget race and more like an infrastructure contest around training data, edge AI, manufacturing, and deployment.
UK’s Under-16 Social Media Plan Pushes Age Checks Into Games and AI Chatbots
The UK is preparing an “Australia plus” under-16 social media plan that would reach beyond TikTok and Instagram into gaming chat, late-night scrolling, AI chatbots, and age-verification systems.
Avalanche’s 11 Million Degree Plasma Is a Compact Fusion Milestone, Not a Power Plant
Avalanche Energy says its compact Jyn fusion device reached apparent ion temperatures above 1 keV, roughly 11 million degrees Celsius. That is a meaningful plasma milestone, but it is still a long way from net power.
State AI Laws Keep Moving as Washington Tries to Stop the Patchwork
States are still advancing AI rules for hiring, lending, chatbot safety, disclosures, and automated decisions even as the White House pushes back against a fragmented regulatory map.
Grok Build Plugin Marketplace Makes Coding Agents a Toolchain Problem
xAI launched a built-in plugin marketplace for Grok Build, turning coding-agent setup into a packaged ecosystem of skills, commands, hooks, MCP servers, and language-server integrations.
Splunk Enterprise Flaw Hits CISA’s Exploited-Vulnerability List
CISA added CVE-2026-20253 to its Known Exploited Vulnerabilities catalog after evidence of active exploitation, and Splunk now says it is aware of limited exploitation. Affected self-managed Splunk Enterprise 10.0 and 10.2 systems should upgrade or carefully apply the PostgreSQL sidecar mitigation.
FCC Burner Phone Proposal Would Turn Phone Privacy Into a KYC Fight
The FCC’s proposed know-your-customer rule would push voice providers to collect and retain more identity data before activating service. The anti-robocall plan also raises a direct fight over prepaid phones, anonymous numbers, and mobile privacy.
Android Fake Call Detection Uses RCS to Fight AI Voice Scams
Google’s Android fake call detection uses an encrypted RCS signal in Phone by Google to warn when a saved contact’s number may be spoofed. The protection is useful, but only works when both phones meet the requirements.
OpenAI Probe Puts ChatGPT’s User Safety Claims Under State Scrutiny
A multistate attorney general investigation is asking for records on ChatGPT safety, advertising, retention, health data, minors, seniors, and model sycophancy. The probe turns consumer AI design choices into a legal and policy test.