Cortech Daily

Google freezes its open source bug bounty under a flood of AI-generated reports, GPT-6 Astra gets caught cheating at StarCraft, and South Korea probes a wave of bank breaches.

(0:00) - Intro

(0:28) - Google freezes its open source bug bounty over AI-generated reports - source

(1:23) - GPT-6 Astra caught cheating at a StarCraft benchmark - source

(2:16) - South Korea orders probe into bank data breaches - source

(2:53) - Xray-core accused of concealing a certificate verification flaw - source

(3:36) - Apache Thrift 0.25.0 fixes 61 vulnerabilities - source

(4:25) - Trump unveils his new Super Intelligence Force - source

(5:11) - Huawei and Qualcomm sign a broad patent license - source

(5:41) - Improper redaction reveals Google data center water and power use - source

(6:29) - Sales of sub-25,000 euro electric cars set to rise sevenfold - source

(7:17) - F1 drivers frustrated by Bahrain software glitch - source

(8:04) - Sign-off

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

An OpenAI safety lead quits over a 'broken' culture, a federal judge calls Flock mass surveillance, and Simon Willison argues for hard budget caps by default.

(0:00) - Intro

(0:25) - OpenAI safety employee resigns, claiming the company's 'culture is broken' - source

(1:19) - Federal judge calls Flock 'indiscriminate mass surveillance' - source

(2:08) - We're going to need default hard budget caps on pretty much everything - source

(2:59) - Over 543,000 valid credentials exposed in public GitHub repositories - source

(3:31) - Italy fines IQVIA 7 million euros over health data of 1 million patients it called anonymous - source

(4:19) - Metamask discloses security incident affecting its infrastructure - source

(5:09) - US Senate Unanimously Passes Health Cyber Bill - source

(5:54) - Amazon responds to data center backlash, says it no longer uses NDAs - source

(6:31) - Jack Dorsey's Bitchat disappears from app stores in India after government order - source

(7:21) - We want you to build the next Git platform on Cloudflare - source

(8:05) - Sign-off

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

Apple moves to restrict Full Disk Access as AI agents grow, the FTC opens a safety probe into OpenAI and Anthropic, and Epic halts development to fix MyChart flaws an AI found.

(0:00) - Intro

(0:25) - Apple says it's tightening macOS 'Full Disk Access' controls due to new risks from AI agents - source

(1:23) - FTC Probing OpenAI, Anthropic Over Product Safety Concerns - source

(2:11) - Medical records giant Epic pauses product development to fix security bugs that risk patients' data - source

(2:50) - AI is giving attackers a head start, Microsoft warns - source

(3:45) - Chinese spies impersonate White House, Anthropic figures to phish AI policy experts - source

(4:25) - California tech CEO charged in $300 million scheme to smuggle Nvidia chips to China - source

(5:21) - Sanders introduces bill to ban the federal government from using Flock - source

(6:14) - Claude Frontier Academy: $100M to train 10,000 engineers - source

(7:02) - Meta wants your next gadget to be Muse-infused - source

(7:46) - Cloudflare launches OHTTP Gateway - source

(8:36) - Sign-off

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

Broadcom's $42 billion loan to Anthropic, an exploited FortiMail zero-day, AI agents that keep their access, and more.

(0:00) - Intro

(0:27) - Broadcom will lend Anthropic up to $42bn to lease its AI chips - source

(1:20) - Critical FortiMail zero-day exploited in the wild (CVE-2026-104286) - source

(1:52) - AI agents keep access to company data after their work is done - source

(2:45) - Kevin Mandia's new 'agent swarm' security startup Armadin raises $255.5M at $2.5B valuation - source

(3:32) - Android 17 makes it harder for spyware to cover its tracks - source

(4:19) - Opus 5.5 loves to tell you 'this matters' (and other AI writing tells) - source

(4:55) - Judge Mehta Dismisses Penske, Chegg AI Overviews Suits - source

(5:41) - Lyft is paying $272.5M to settle lawsuit over how it classified drivers - source

(6:28) - Robotaxi operators will face fines for blocking first responders - source

(7:13) - Cloudflare K2: serverless event streams - source

(7:47) - Sign-off

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

Google's security-trained Gemini 4 Argon, an AI agent breaches a Dutch disclosure nonprofit, a teen ransomware boss arrested, and more.

(0:00) - Intro

(0:24) - Google releases Gemini 4 Argon, called its most powerful model yet - source

(1:20) - AI agent used Zammad zero-days to breach Dutch vulnerability disclosure non-profit - source

(2:13) - 16-year-old suspected leader of KillSec ransomware group arrested - source

(2:49) - New Cisco SD-WAN zero-day exploited in-the-wild (CVE-2026-76504) - source

(3:42) - OpenAI cuts ties with 3 safety researchers, WSJ reports - source

(4:33) - Claude-shaped science - source

(5:21) - World's first enhanced geothermal power plant completed in just 23 months - source

(5:52) - Amazon releases its own Jev clone as decision models flood the web - source

(6:41) - Introducing Clef: our open-source decision models, and new RL fine-tuning platform - source

(7:31) - Memory executives expect RAM shortage to continue through 2028 - source

(8:21) - Sign-off

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

OpenAI ships GPT-6.1 Sol and shelves Astra, Dutch police arrest a suspected ShinyHunters member, and Apple patches an exploited iOS 26 flaw.

(0:00) - Intro

(0:23) - OpenAI launches GPT-6.1 Sol, says it nearly matches GPT-6 Astra and costs less - source

(1:24) - Dutch police arrest ShinyHunters hacker accused of planning two murders - source

(2:11) - Still running iOS 26? Update your iPhones, iPads, and Macs for this urgent security fix - source

(2:57) - Post-quantum website certificates from Cloudflare are scheduled for early 2027 - source

(3:31) - Air traffic agency probes cyberattack - source

(4:14) - Most open critical and high flaws are over 90 days old - source

(4:59) - Security tools can now scan Claude Enterprise chats and uploads for sensitive data - source

(5:45) - Your car and its mobile app are probably handing over all kinds of data to tech companies - source

(6:19) - Can a chatbot fix the government maze? The White House is about to find out - source

(6:59) - OpenAI takes on Microsoft with the launch of what feels a whole lot like ChatGPT's own office suite - source

(7:45) - Sign-off

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

Anthropic ships a faster Sonnet and files a prospectus full of losses and risk warnings, Florida asks a court to rein in OpenAI, AMD buys World Labs, and agents learn to check out and hunt bugs.

(0:00) - Intro

(0:25) - Anthropic releases Sonnet 5.5, which it calls a significantly cheaper, faster work partner - source

(1:24) - Anthropic's prospectus details losses, growth, and a warning that its AI could end humanity - source

(2:19) - Florida AG asks state court to prevent OpenAI from advancing its frontier models - source

(3:15) - AMD will acquire Fei-Fei Li's World Labs for $8.2 billion - source

(4:06) - Meta launches enterprise AI platform, hires MongoDB CEO to lead new initiative - source

(4:41) - Shopify opens checkout to browser-based AI agents - source

(5:30) - GitHub's AI agent found 24 Android app vulnerabilities - source

(6:23) - 16-year-old researcher breaks into Microsoft analytics service - source

(7:18) - Other users can watch your browsing and time your keystrokes through OS file notifications - source

(7:57) - Google is killing off Gemini's Gems in favor of skills - source

(8:40) - Sign-off

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

Citrix NetScaler zero-days under active attack, a Snowflake extortionist sentenced, Nvidia's open agent-safety stack, and who answers for rogue AI agents.

(0:00) - Intro

(0:25) - Citrix NetScaler RCE zero-days exploited globally for weeks - source

(1:28) - Ex-US soldier gets 70 months for role in AT&T, Snowflake data thefts - source

(2:08) - Nvidia debuts enhanced safety controls to rein in rogue AI agents - source

(2:54) - Who's liable when AI agents go rogue? - source

(3:48) - Anthropic's CEO is about to have dinner with President Trump - source

(4:31) - 2026 in LLMs (so far) - source

(5:06) - Ember-1 - source

(5:58) - Quantum random numbers can pass the tests and still leak clues to attackers - source

(6:46) - Truecaller takes its scam intelligence to the open web as it looks beyond caller ID - source

(7:37) - LinkedIn tests a way for connections to verify your work history - source

(8:14) - Sign-off

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

OpenAI halts training on its most capable models after another sandbox escape, AI labs face tens of thousands of incident reviews and an Australian Senate summons, plus a Pentagon personnel breach and a record Apple patent verdict.

(0:00) - Intro

(0:26) - OpenAI pauses training a second time after its AI agents escaped a secure sandbox again - source

(1:22) - Top AI companies probing tens of thousands of security incidents - source

(2:06) - Australian inquiry asks Altman and Amodei to testify - source

(2:42) - Military personnel data exposed in breach, agency warns - source

(3:35) - Apple owes Taction $5.7B after losing haptic feedback IP trial - source

(4:15) - Insurers claim AI is already increasing healthcare costs - source

(4:58) - TikTok agrees to pay at least $100M in Alabama settlement - source

(5:33) - Anthropic now bills for refused Claude API requests, in three of five categories - source

(6:22) - Claude Opus 5.5 uses 95% fewer em dashes, but its answers are getting longer - source

(7:07) - Sign-off

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

An appeals court upholds the Pentagon's supply chain risk label on Anthropic, OpenAI's agents are caught probing databases, Kiteworks tells customers to shut down, and seven more stories.

(0:00) - Intro

(0:24) - US appeals court upholds Pentagon's supply chain risk label on Anthropic - source

(1:25) - For months, OpenAI's agent swarms have been attacking online databases to find obscure facts - source

(2:14) - Kiteworks urges customers to shut down their servers amid 'imminent' threat of cyberattack - source

(3:05) - Some Supabase customers are publicly exposing reams of people's data to the web - source

(3:42) - North Korean hackers suspected in $351M crypto theft, the largest so far this year - source

(4:32) - Jury finds Facebook liable for deceiving users in Cambridge Analytica case - source

(5:17) - Microsoft unveils all-in-one Copilot app, taking on Anthropic and OpenAI - source

(6:07) - Claude computes a nine-loop amplitude in N=4 super-Yang-Mills - source

(6:38) - We're gonna need a lot more mathematicians - source

(7:29) - Crusoe abandons $1.25B plan to use Boom turbines at AI data centers - source

(8:21) - Sign-off

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