Homechevron_rightArchives

Archives

Explore Stories by Date

30 Mar 2026

Apple Intelligence accidentally goes live in China, then disappears without regulatory approval

Apple Intelligence briefly appeared on iPhones across mainland China in the early hours of Tuesday morning before being pulled offline. The accidental launch occurred without regulatory approval from China's Cyberspace Administration, potentially exposing Apple to administrative penalties. Bloomberg's Mark Gurman confirmed the rollout was an error, noting Apple would never launch AI features in its most important international market without announcement.

6 Sources

OpenAI patches critical ChatGPT and Codex security vulnerabilities exposing user data and tokens

OpenAI addressed two critical security vulnerabilities in February 2026. Check Point discovered a ChatGPT flaw enabling silent data exfiltration through DNS tunneling, bypassing AI guardrails without user consent. BeyondTrust found a command injection vulnerability in Codex allowing GitHub token theft through malicious branch names, potentially compromising enterprise organizations.

6 Sources

Micron develops stacked GDDR memory as cost-effective HBM alternative for AI inference

Micron is developing a vertically stacked GDDR memory architecture positioned between traditional GDDR and high-bandwidth memory (HBM). The company aims to introduce early prototypes by 2027, targeting AI inference workloads that prioritize memory capacity and cost-efficiency over peak bandwidth. This innovation could reshape memory solutions for GPUs and AI accelerators.

2 Sources

Samsung's AI Wine Fridge Scans Labels and Tracks Your Collection, But Critics Question the Value

Samsung launched the Infinite AI Wine Refrigerator in South Korea, featuring an AI Vision camera that scans wine bottle labels to automatically track inventory. Priced at $4,300, the fridge stores 101 bottles across three cooling zones. Wine experts and tech reviewers question whether the AI features justify the premium price, with some calling it another example of companies adding AI branding without substantial functionality.

3 Sources

Google Maps brings AI-powered EV trip planning to Android Auto for 350+ electric vehicles

Google Maps now offers AI-powered EV trip planning on Android Auto, addressing range anxiety for electric vehicle owners. The feature uses advanced energy models to predict battery consumption across 350+ EV models from 16 brands, including Audi, BMW, Kia, and Mercedes-Benz, while factoring in real-time traffic, road elevation, and weather conditions.

9 Sources

Microsoft Copilot now uses GPT and Claude AI models together to fact-check research outputs

Microsoft unveiled a multi-model approach for its Copilot Researcher tool that leverages both OpenAI's GPT and Anthropic's Claude simultaneously. The new Critique feature has GPT draft responses while Claude reviews for accuracy, achieving a 13.8% improvement on industry benchmarks. The company also rolled out Copilot Cowork to early-access customers as competition intensifies in enterprise AI.

12 Sources

ScaleOps raises $130M to tackle computing efficiency crisis as AI infrastructure costs soar

ScaleOps has raised $130 million in Series C funding at an $800 million valuation to address the growing problem of wasted compute resources in AI infrastructure. The startup's platform autonomously manages Kubernetes infrastructure in real-time, reducing cloud and AI infrastructure costs by up to 80%. Led by Insight Partners, the round brings total funding to $210 million as the company reports 450% year-over-year growth.

3 Sources

Tennessee Grandmother Wrongly Jailed for Five Months After AI Facial Recognition Error

Angela Lipps, a 50-year-old Tennessee grandmother, spent over five months in jail after Clearview AI's facial recognition system mistakenly linked her to bank fraud cases in North Dakota. Despite never visiting the state, she was arrested, extradited, and held until bank records proved her innocence. The case highlights serious concerns about law enforcement's reliance on AI technology without proper corroboration.

4 Sources

AI Smart Glasses Fuel Exam Cheating in China as Students Rent Devices for $6 a Day

A rental market for AI smart glasses has emerged in China, where students pay $6 to $12 daily to cheat on exams. Devices from Rokid and Meta scan test questions and project real-time answers onto lenses, bypassing detection due to their discreet design. Despite explicit bans, enforcement remains limited as the technology becomes harder to distinguish from ordinary eyewear.

4 Sources

AI Agent Calls 3,000 Irish Pubs to Track Guinness Prices, Drives Down Costs Across Ireland

Matt Cortland built an AI voice agent named Rachel that called over 3,000 Irish pubs to track the cost of Guinness after paying €7.80 for a pint in Dublin. The resulting Guinndex consumer price index revealed an average price of €6.01, exposing wide price variations. Pubs are now lowering their prices to compete, with some dropping costs by €0.40.

3 Sources

Rebellions raises $400 million at $2.34 billion valuation as AI chip startup eyes US expansion

South Korean AI chip startup Rebellions has secured $400 million in a pre-IPO funding round led by Mirae Asset Financial Group and Korea National Growth Fund, valuing the company at $2.34 billion. The company has raised $650 million in six months and launched RebelRack and RebelPOD infrastructure platforms while targeting Meta and xAI as US customers.

7 Sources

Starcloud raises $170 million to build AI data centers in space, reaching unicorn status

Starcloud has raised $170 million at a $1.1 billion valuation to build orbital data centers powered by solar energy. The startup already launched the first Nvidia H100 chip in space and demonstrated AI model training in orbit. But the business model depends on unproven technology and SpaceX's Starship rocket, which isn't flying commercially yet.

10 Sources

OpenClaw Security Risks Escalate as Critical Vulnerability Exposes Thousands of AI Agent Instances

A critical vulnerability in OpenClaw, the viral AI agent tool with 347,000 GitHub stars, has security experts urging users to assume compromise. CVE-2026-33579 allows attackers with minimal permissions to gain full administrative control, potentially exposing sensitive data across thousands of unprotected instances. The incident highlights the inherent security risks of autonomous AI agent operations.

7 Sources

Australian man uses ChatGPT to design AI vaccine for his dog's cancer, catches Sam Altman's eye

Sydney-based AI consultant Paul Conyngham turned to ChatGPT, AlphaFold, and other AI tools to create a personalized experimental treatment for his rescue dog Rosie's terminal mast cell cancer. After months of AI-driven research and collaboration with university scientists, Rosie achieved partial remission with her largest tumor shrinking dramatically. OpenAI CEO Sam Altman praised the effort as an amazing story, though experts emphasize the need for rigorous scientific validation.

3 Sources

DeepSeek suffers 7-hour outage, its longest service disruption since R1 and V3 models launched

China's DeepSeek AI chatbot went offline for over 7 hours on March 30, marking the longest outage since its viral debut in early 2025. The disruption affected the platform's web interface, which had maintained near-perfect uptime for over a year. The incident raises questions about platform reliability as the company prepares to unveil its anticipated V4 model.

4 Sources

Get all updates into your inbox 💌, Stay Up-to-date for Free 🥳

Your Daily Dose of Curated AI News

Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Spend less time searching for the latest in AI and get straight to action.

The Outpost

News

About