AI News Archive - Browse Page 97 of 182
Browse AI news articles covering LLMs, tools, research, and industry trends
Oz report details why firms fail, teams save hours, 60% of agent PRs succeed
Most companies are failing to build useful AI agents. A new report from Oz explains why, and the numbers are stark.
Samsung adds Perplexity AI agent to Galaxy, enabling OS integration
Samsung's next flagship feature isn't a camera. It’s a calculated gamble that you’ll want a different AI than its own.
Early 2026: Privacy Helplessness Grows as Meta Tests Limits in US
By early 2026, the quiet acceptance has already settled in. Americans scroll past privacy warnings with a numb shrug, convinced the fight is over...
Test Shows ‘-ai’ Trick Blocks Google AI Overviews Only on Desktop Browsers
Want to see the actual internet? On a desktop browser, just add “-ai” to your Google search. The AI Overview vanishes. It’s a clean, instant hack.
Sampath's approach: Teams stitch agents, models, systems for future work
Forget the single killer app. The future of your company will be assembled, a patchwork of agents, models, and old systems crudely wired together...
Tumbler Ridge shooter shared scenarios with ChatGPT; OpenAI did not notify police
Months before the shootings at Tumbler Ridge Secondary, Miles Rootselaar typed his violent fantasies into ChatGPT. OpenAI’s safety teams saw them.
Google Gemini 3.1 Pro doubles reasoning performance in benchmark
Google's latest AI model doesn't just edge ahead—it vaults. On the ARC-AGI-2 benchmark, a test of novel abstract thought, Gemini 3.1 Pro scored...
Runlayer adds OpenClaw security, boosting prompt injection resistance to 95%
Runlayer built a bouncer. For your AI agents. The startup’s new security features for OpenClaw just turned a theoretical nightmare into a quantified...
Microsoft Copilot ignored sensitivity labels twice; DLP missed both
The security stack reported all-clear, because it never saw the layer where the violation occurred.
Blue Oval kills F‑150 Lightning, GM books USD 7.6 B charge, Stellantis follows
Ford didn't just pause the F-150 Lightning. They killed it. That one decision triggered a cascade of financial panic across Detroit, revealing an...
Trump Mobile unveiled as gold‑foil Liberty Mobile, termed “enabler”
The phone is a stage prop. The real machinery, the licensing, the tax compliance, the E911 fees, the carrier negotiations, belongs to a company...
Pentagon says AI firms must partner despite concerns over war‑crime claims
The Pentagon is hunting in Silicon Valley, but not for chatbots. It wants AI built for a single, stark purpose: winning wars.
AWS blames staff after Kiro AI coding assistant triggers 13‑hour outage
For thirteen hours last December, an AWS system in mainland China was down. According to the Financial Times, the culprit was Amazon's own Kiro AI...
Handshake refusal highlights strained ties among leading AI labs
The internet loves a snub. A hand offered and ignored makes for perfect, shareable cringe. That frozen moment from the AI Summit?
NVIDIA Co-Design Boosts Sarvam AI Inference, Cuts TTFT Below One Second
The race to sub-second time-to-first-token is a brutal one. For Sarvam AI’s sovereign models, the baseline was functional but far from fast enough.
Run:ai on 64 GPUs serves 10,200 users, matching native scheduler
Forget the hype. NVIDIA's Run:ai just ran a brutally simple test to see what happens when you actually use the thing: take 64 GPUs and throw users at...
The Pitt examines why doctors might embrace generative AI in clinical practice
The medical profession has two rules for new technology. First, be late. Second, be very, very sure. For decades this worked.
Perplexity abandons ads as investors once eyed billions of users
The dream was a Google killer. The reality, according to third-party analytics firm Similarweb, is a search engine with fewer monthly users than the...
Hacker Exploits Cline AI Coding Agent Vulnerability Highlighted by Researcher
A security researcher points out a flaw. Three days later, a hacker uses it. This is now the standard timeline for AI vulnerabilities.
Anthropic's mid-tier model offers 30‑minute ChatGPT crash course, 100+ prompts
Thirty minutes. That’s all it takes to go from ChatGPT novice to someone who actually gets things done with AI.