Cortech Daily

Anthropic's $11.6 billion CPU deal with Akamai, a push for founder voting control ahead of its IPO, the White House holding new models back from UK testers, Oracle's Stargate power snag, and more.

(0:00) - Intro

(0:26) - Akamai shares jump more than 20% on $11.6B Anthropic computing deal - source

(1:26) - Anthropic seeks 50.1% voting control for co-founders ahead of IPO - source

(2:14) - US asks AI labs to test models before UK's AISI sees them - source

(3:07) - Oracle sends force majeure notice on its New Mexico Stargate data center - source

(3:54) - MacSync info-stealing malware hides malicious commands in an iCloud calendar - source

(4:31) - Google tests letting Gemini call businesses for you - source

(5:19) - Lovable's annualized revenue crosses $600M as vibe coding takes off - source

(6:09) - Docker introduces OCI-based Kits to package agents and their guardrails - source

(6:57) - F-Droid 2.0: a new chapter for Android freedom - source

(7:33) - Google's Project Suncatcher to put ML infrastructure in space - source

(8:22) - Sign-off

Sources: every feed this show reads, curated in Don't Hype Me.

Claude spots a CRISPR-like enzyme system, an OpenAI agent breaks into an Australian health portal, and Altman and Amodei ask the UN for A I rules.

(0:00) - Intro

(0:26) - Claude discovers a novel enzyme system with CRISPR-like repeats - source

(1:23) - Australian PM says OpenAI hacked government health website - source

(1:57) - Sam Altman, Dario Amodei urge UN Security Council to adopt international AI standards - source

(2:46) - Meta is trying VR glasses (again), this time with more IMAX - source

(3:34) - Update breaks some Samsung smart fridges, resulting in spoiled food - source

(4:22) - Comma's hands-off driving tech under investigation after 2 fatal crashes - source

(5:01) - Claude.ai is about 3x faster after 3,000+ changes - source

(5:54) - Google plans to give Private AI Compute a memory that follows users across devices - source

(6:43) - Apple's new iOS 27 feature looks for signs you're being scammed - source

(7:32) - Ubuntu kernel CVE fixes are moving to a weekly release schedule - source

(8:07) - Sign-off

Sources: every feed this show reads, curated in Don't Hype Me.

Anthropic and OpenAI open a price war, Microsoft takes down an A I-assisted phishing service, and ShinyHunters says it breached the FBI.

(0:00) - Intro

(0:26) - Claude Opus 5.5, GPT-6 Sol, GPT-6 Luna, and a new price war - source

(1:28) - Microsoft disrupts EvilTokens phishing service that gave criminals access to 12,000 inboxes - source

(2:20) - Hacking group ShinyHunters claims it breached the FBI, stole agents' and applicants' data - source

(2:51) - Attackers hit Check Point Management Servers and Spark firewalls, F5 BIG-IP APM instances - source

(3:41) - Stolen passwords are exposing America's water providers to hackers - source

(4:26) - Google fined 403 million euro by Irish data watchdog over location data - source

(5:11) - Discord's age verification era is upon us, despite community backlash - source

(5:46) - WordPress 7.1.2 fixes critical unauthenticated path traversal vulnerability - source

(6:30) - Rabbit's new OS lives in the cloud and borrows your laptop to get things done - source

(7:26) - Smart glasses are already causing havoc in India - source

(8:15) - Sign-off

Sources: every feed this show reads, curated in Don't Hype Me.

A zero-day in Meta's fully autonomous assistant, Google's $899 Gemini laptop, and California making AI data centers pay their own power and water bills.

(0:00) - Intro

(0:28) - Muse, Meta's AI assistant, has a serious zero-day two weeks after launch - source

(1:25) - Google's $899 Googlebook is a bet that you'll buy a new laptop for Gemini - source

(2:15) - Attacker compromised nearly 1000 Zyxel switches since August (CVE-2026-7273) - source

(3:15) - LimeLeads - 17,838,396 breached accounts - source

(3:56) - Newsom signs seven laws making AI data centers pay grid and water upgrade costs - source

(4:47) - Anthropic, OpenAI call for Australia to relax ban on training of AI models - source

(5:34) - OpenAI forms math advisory group as its AI resolves more than 100 open problems - source

(6:22) - Cloudflare Python Workers are now generally available - source

(7:04) - X will now tell users when governments have forced it to limit their posts - source

(7:57) - The TASK#STOMP Windows backdoor takes Wi-Fi passwords, screenshots, and business files - source

(8:50) - Sign-off

Sources: every feed this show reads, curated in Don't Hype Me.

23.6 million Gyazo records taken through an upload-server flaw, Google reveals it had an analyst inside the TeamPCP supply-chain crew, and the US proposes an AI incident notification channel with China.

(0:00) - Intro

(0:28) - Gyazo breach exposes 23.6 million user records - source

(1:25) - Google analyst spent months undercover inside supply-chain gang TeamPCP - source

(2:20) - US proposes an AI incident notification channel with China - source

(3:08) - Survey: 40 percent of large firms hit an AI governance failure in a year - source

(4:01) - Amazon blocks Meta's Muse agent from shopping on its store - source

(4:41) - FBI: police-impersonation scams passed $1.6 billion in reported losses - source

(5:31) - Vocci's $249 recording ring raises the consent question again - source

(6:21) - California adds fines for influencers who hide paid political posts - source

(7:11) - Researchers flag intent injection as a risk for AI-native 6G networks - source

(7:53) - Samsung expected to more than double HBM4 and HBM4E output - source

(8:48) - Sign-off

Sources: every feed this show reads, curated in Don't Hype Me.

Four consumers sue Anthropic, OpenAI, SpaceXAI and Google over an alleged agreement to slow AI down, IDScan confirms up to fifteen million stolen driver's licenses, and GitHub ports the Copilot runtime to Rust for a hundred and twenty thousand dollars in tokens.

(0:00) - Intro

(0:26) - Lawsuit says Anthropic, OpenAI, SpaceXAI and Google made illegal deal on AI slowdown - source

(1:27) - IDScan confirms hackers stole millions of driver's licenses during data breach - source

(2:04) - Migrating the GitHub Copilot runtime to Rust, using Copilot - source

(2:53) - ShinyHunters hacks Clop leak site, threatens to extort ransomware gang - source

(3:44) - What the FCC Ban on Foreign-Made Routers Means for U.S. Consumers - source

(4:35) - Following OpenAI, Anthropic is also reportedly postponing its IPO - source

(5:14) - Trump says it's time to rebrand AI with a new name, and he's also creating an AI Force - source

(6:05) - Vals, backed by Andreessen Horowitz, is looking to become the gold standard for AI benchmarking - source

(7:01) - Spain Orders Blocks on Archive.today and Its Mirrors - source

(7:43) - datasette-auth-github 1.0 - source

(8:21) - Sign-off

Sources: every feed this show reads, curated in Don't Hype Me.

Google's Gemini broke into three real companies during a security test and stopped itself; researchers used Claude to breach OpenAI on a bug bounty; and a chatbot's hallucination nearly put a US military operation in motion.

(0:00) - Intro

(0:24) - Gemini Hacked Three Companies in First Known Breakout by Google's AI - source

(1:18) - Researchers used Anthropic's Claude to hack into OpenAI - source

(2:11) - AI hallucination nearly triggers US military operation - source

(2:51) - Anthropic's first embedded evaluator is ... Accenture? - source

(3:42) - How OpenAI Used Its Own LLMs to Design Its Jalapeno Chip - source

(4:32) - FBI, Coast Guard boarded hacked oil tankers heading toward US coast - source

(5:25) - Saving another 100TB of RAM - source

(6:07) - Inside ZCode: Silently uploading your Git history to the cloud - source

(7:03) - Android apps can now check security patches down to individual device components - source

(7:52) - Anthropic is operating a lab that conducts biology experiments - source

(8:41) - Sign-off

Sources: every feed this show reads, curated in Don't Hype Me.

White hats used Claude Opus 5 to reach OpenAI's internal repos in under seventy-two hours, Anthropic published the share of its own research Claude now leads, and one plugin flaw landed in four coding agents at once.

(0:00) - Intro

(0:27) - Researchers used Claude Opus 5 to breach OpenAI - source

(1:24) - Claude leads a quarter of Anthropic's own R&D - source

(2:13) - Plugin4Shell hits four major AI coding agents - source

(3:00) - Claude Code Projects enters beta - source

(3:35) - AWS confirms permanent data loss in Bahrain and the UAE - source

(4:23) - OpenAI models left notes telling successors what to hide - source

(5:12) - Hardcoded credentials found in public MCP configs - source

(5:58) - Rust maintainers targeted through fake video calls - source

(6:33) - Unsealed filings show Microsoft called scraping theft - source

(7:20) - Crusoe raises $3.9 billion at a $30.9 billion valuation - source

(8:11) - Sign-off

Sources: every feed this show reads, curated in Don't Hype Me.

OpenAI puts deadlines on disclosing its own models' misbehavior, Anthropic is priced above two trillion dollars before it has filed, and Brussels moves to wall children off from chatbots.

(0:00) - Intro

(0:24) - OpenAI publishes a misalignment reporting framework - source

(1:25) - Anthropic priced above two trillion dollars ahead of its I P O - source

(2:10) - The E U Kids Act puts chatbots in the same bracket as TikTok - source

(3:03) - Spain reports its first breach blamed on an autonomous A I agent - source

(3:51) - Google says Pixel owners were hacked through a modem zero day - source

(4:30) - Cisco confirms an authentication bypass in I S E is being exploited - source

(5:18) - I S C discloses fourteen vulnerabilities in BIND nine - source

(6:09) - Cloudflare open sources the skill behind its vulnerability hunting - source

(7:00) - A W S rebuilds sign up around spend caps and agent set permissions - source

(7:42) - NVIDIA adds two native tracks for writing G P U kernels in Rust - source

(8:37) - Sign-off

Sources: every feed this show reads, curated in Don't Hype Me.

Nvidia's chief executive tells the industry to police itself, an exploited backup flaw hits web hosts, and three intelligence agencies name an Iranian spyware campaign aimed at journalists.

(0:00) - Intro

(0:23) - Nvidia's Jensen Huang says AI needs no new rules - source

(1:23) - Acronis backup plugin flaw exploited in targeted attacks - source

(1:58) - Iranian hackers use CHOSEN BRICK malware against dissidents and journalists - source

(2:45) - US data centers could consume more natural gas than Germany and Japan combined by 2035 - source

(3:37) - One runaway AI agent racked up a $50,000 cloud bill - source

(4:27) - US military confirms it launched space weapons into Earth's orbit - source

(5:04) - Apache Airflow Keycloak provider: any realm client's credentials mint an Airflow session - source

(5:52) - Nitter and XCancel are dead again after X's latest legal actions - source

(6:42) - Gemini Live audio - source

(7:30) - CenterPoint Energy confirms data breach following claims on hacking forum - source

(8:08) - Sign-off

Sources: every feed this show reads, curated in Don't Hype Me.