ax@ax-radar:~/feed $ tail -f signal.log
41 srcsignal 1208%cycle 04:32

hot events · 2026-06-15

4 signals · updated 3m ago
live · 217 today·policy v2
LATENT SPACEAnthropic pulls Fable and Mythos after US e…96·LATENT SPACEAnthropic launches Claude Fable 5, its firs…88·HACKER NEWS FRONTPAGDid Anthropic ask for its own export contro…82·HACKER NEWS FRONTPAGAnthropic flies senior technical staff to D…82·AI HOT (CURATED POOLWSJ: OpenAI weighs steep price cuts and pla…82·HACKER NEWS FRONTPAGBram Cohen: Claude is turning into an assho…78·R/LOCALLLAMAXiaomi serves MiMo V2.5 at 1000–3000 tps wi…78·IMPORT AI (JACK CLARAI learns to game society's rules, and Anth…78·MIT TECHNOLOGY REVIEGoogle DeepMind is worried about what happe…78·DWARKESH PATELThe sample efficiency black hole: AI models…78·LATENT SPACECognition launches FrontierCode: a coding b…78·HACKER NEWS FRONTPAGGabriel Weinberg argues with data that “eve…78·LATENT SPACEAnthropic pulls Fable and Mythos after US e…96·LATENT SPACEAnthropic launches Claude Fable 5, its firs…88·HACKER NEWS FRONTPAGDid Anthropic ask for its own export contro…82·HACKER NEWS FRONTPAGAnthropic flies senior technical staff to D…82·AI HOT (CURATED POOLWSJ: OpenAI weighs steep price cuts and pla…82·HACKER NEWS FRONTPAGBram Cohen: Claude is turning into an assho…78·R/LOCALLLAMAXiaomi serves MiMo V2.5 at 1000–3000 tps wi…78·IMPORT AI (JACK CLARAI learns to game society's rules, and Anth…78·MIT TECHNOLOGY REVIEGoogle DeepMind is worried about what happe…78·DWARKESH PATELThe sample efficiency black hole: AI models…78·LATENT SPACECognition launches FrontierCode: a coding b…78·HACKER NEWS FRONTPAGGabriel Weinberg argues with data that “eve…78·LATENT SPACEAnthropic pulls Fable and Mythos after US e…96·LATENT SPACEAnthropic launches Claude Fable 5, its firs…88·HACKER NEWS FRONTPAGDid Anthropic ask for its own export contro…82·HACKER NEWS FRONTPAGAnthropic flies senior technical staff to D…82·AI HOT (CURATED POOLWSJ: OpenAI weighs steep price cuts and pla…82·HACKER NEWS FRONTPAGBram Cohen: Claude is turning into an assho…78·R/LOCALLLAMAXiaomi serves MiMo V2.5 at 1000–3000 tps wi…78·IMPORT AI (JACK CLARAI learns to game society's rules, and Anth…78·MIT TECHNOLOGY REVIEGoogle DeepMind is worried about what happe…78·DWARKESH PATELThe sample efficiency black hole: AI models…78·LATENT SPACECognition launches FrontierCode: a coding b…78·HACKER NEWS FRONTPAGGabriel Weinberg argues with data that “eve…78·
RSS live
2026-06-15 · Mon
00:07
10h ago
STILL DEVELOPING · 1d● P1New York Times Chinese· rssZH00:07 · 06·15
Google sues China-based cybercrime ring for using Gemini to mass-produce fake sites
Google filed a lawsuit against a China-based cybercrime ring called Outsider Enterprise, accusing it of using Gemini to build 131 software toolkits that mass-produce fake sites impersonating Google, USPS, and E-ZPass. In just two weeks this May, the group sent 2.5 million phishing texts to Android users, linking to 9,000 fake sites. Google says this is its first coordinated takedown with the FBI and carriers AT&T, T-Mobile, and Verizon. The FBI reported roughly $893 million in AI-linked fraud losses last year; Google estimates hundreds of thousands of victims here and millions of dollars in losses. The post does not name specific defendants or their locations.
#Code#Google#Gemini#FBI
why featured
Google's first legal action against AI-enabled fraud rings, backed by concrete numbers and cross-border coordination. Capped below 85 because it's ultimately a law-enforcement story, not an AI capability or product update.
editor take
Google sued a Chinese scam group for using AI to blast 2.5M phishing texts; all three sources agree because they're working from Google's own court filing, so the facts are solid.
sharp
This one's worth opening because Google handed the court filing to reporters — TechCrunch and two Chinese tech outlets are all working from the same document, so there's no independent reporting divergence. The core numbers: a group called Outsider Enterprise used AI to generate and send 2.5 million scam texts over two weeks, claiming hundreds of thousands of victims. Google filed a civil suit in US federal court, aiming to shut down their infrastructure through legal means. Where I'd discount: "hundreds of thousands of victims" is Google's own estimate, not a third-party count — the real number could be lower or higher. Also, the filing doesn't say what AI tools the group used. Could be a homegrown chat model, could be a wrapper around a public API. We just don't know. The thing to actually watch isn't the AI angle — it's Google choosing litigation over purely technical countermeasures for cross-border fraud. Whether a US civil judgment can reach defendants in China is a big open question.
HKR breakdown
hook knowledge resonance
open source
92
SCORE
H1·K1·R1

more

feeds

admin