--- title: "🎙️ Pentagon asks OpenAI to remove refusal button? IDScan leaks 150M driver's licenses | AI Daily Podcast" description: "The Pentagon actually asked OpenAI to help customize an AI system that \"won't say no\"—could this be the starting point of out-of-control technology from sci-fi movies? Today on *Mark's Tech Insights*, host Mu Yan will take you through this military contract that has shocked the tech world. Plus, we'll talk about IDScan, a company you've probably never heard of that quietly leaked 150 million US and Canadian driver's license records, and what god-tier technology DeepMind used to calculate all 9 billion human genetic mutations at once." canonical_url: "https://blog.markkulab.net/en/tech-news/ai-daily-podcast-2026-09-11" author: "Mark Ku" author_url: "https://blog.markkulab.net/en/author/mark-ku" site: "Mark Ku's Tech Notes" date_published: "2026-09-11 10:00:00 +0800" category: "What's New in Tech" tags: ["ai-daily", "podcast", "tech-news", "Mark的TechInsights", "AI日報", "OpenAI", "DeepMind", "Suno", "IDScan", "NVIDIA", "個資安全"] language: "en" license: "CC BY 4.0" license_url: "https://creativecommons.org/licenses/by/4.0/" attribution: "when reusing or quoting, credit the author and link back to the original" --- # 🎙️ Pentagon asks OpenAI to remove refusal button? IDScan leaks 150M driver's licenses | AI Daily Podcast ## Introduction The Pentagon has actually asked OpenAI to help customize an AI system that "won't say no." Could this be the starting point of out-of-control technology straight out of a sci-fi movie? Today on *Mark's Tech Insights*, your host, Muyan, will take you through this military contract that has shocked the tech world. In addition, we'll talk about IDScan—a company you've probably never heard of that quietly leaked the personal data of 150 million US and Canadian driver's licenses—and the god-tier technology DeepMind used to calculate all 9 billion human genetic mutations in one go. Stay tuned for these exciting stories coming up next! ## Today's Featured News ### 1. IDScan Confirms 150 Million US and Canadian Driver's Licenses Stolen, Personal Data Sold on Dark Web - **Source**: TechCrunch (） - **Summary**: Identity verification giant IDScan confirmed that its servers were breached by hackers, resulting in the leak of over 150 million US and Canadian driver's licenses and passport data, which are now listed for sale on the dark web. The company primarily provides ID scanning and age verification services for venues like nightclubs and dispensaries. - **Most Surprising Point**: The entity that actually leaked your ID isn't the big brand you trust, but rather the third-party verification vendor they outsourced to—one whose name you've probably never even heard of. - **Taiwan Perspective**: This serves as a wake-up call for Taiwan's fintech and digital identity verification (KYC) sectors. Many Taiwanese startups and banks also outsource ID verification to third-party vendors. If a leak occurs, the brand's reputation will suffer a devastating blow. - **Discussion Points**: 1. In our daily lives, how many "invisible third parties" are we actually handing our personal data over to? 2. How should companies implement stricter cybersecurity audits when choosing outsourced KYC vendors? - **Script Suggestion**: Have you ever wondered where your personal data actually goes when a bouncer or clerk scans your ID at a nightclub or store? IDScan, the company at the center of this massive breach, is the behind-the-scenes vendor that helps these businesses scan IDs and verify ages. As it turns out, they were quietly hacked, and 150 million driver's licenses and passports were put up for sale on the dark web! The scariest part is that most of these victims have never even heard of this company, let alone known that their personal data had been passed along. It's like giving your keys to your building manager, who then quietly takes them to an unknown locksmith to make copies. Eventually, the locksmith's shop gets robbed, and your keys are compromised. This kind of "invisible third-party" security vulnerability is incredibly hard to guard against! --- ### 2. Evading Merger Review? US DOJ Launches Antitrust Investigation into NVIDIA's $20 Billion Deal with Groq - **Source**: Bloomberg (） - **Summary**: The US Department of Justice is investigating a massive $20 billion deal between NVIDIA and AI chip startup Groq. The DOJ suspects that NVIDIA deliberately packaged a "de facto acquisition" as a "licensing agreement" to evade antitrust review, as the deal resulted in Groq's founders and almost its entire core team joining NVIDIA. - **Most Surprising Point**: Through the clever contractual trick of "buying the people, not the company," NVIDIA successfully bypassed the entire merger review system. - **Taiwan Perspective**: NVIDIA is constantly embroiled in antitrust controversies. For Taiwan's semiconductor supply chain, any move by such a giant could affect capacity allocation. Taiwanese foundries need to closely monitor how this lawsuit impacts the competitive structure of the chip market. - **Discussion Points**: 1. Will this "buying the people, not the shell" contracting method become the new normal for tech giants to evade regulation? 2. Will the DOJ's intervention this time effectively curb NVIDIA's absolute monopoly in the AI chip sector? - **Script Suggestion**: This move is absolutely brilliant! NVIDIA spent $20 billion to partner with Groq, framing it on the surface as a "technology licensing" deal where Groq remains an independent company. But in reality, Groq's founders and most of their core R&D team have already packed up and gone to work for NVIDIA! This is essentially an acquisition disguised as a "golden shell escape." The DOJ has caught on and believes NVIDIA is exploiting contract loopholes to bypass antitrust reviews. If this tactic works, any tech giant looking to swallow a small startup in the future could just buy up all the talent and leave an empty shell behind. Wouldn't that make antitrust laws completely useless? --- ### 3. Pentagon Splurges Hundreds of Millions, Demanding OpenAI Build a Custom Military AI That "Won't Refuse Commands" - **Source**: The Intercept (） - **Summary**: According to newly exposed US Department of Defense contracts, giants like OpenAI and Anthropic have signed deals with the military worth up to $200 million. The Pentagon explicitly requested that OpenAI provide a customized version that "minimizes refusals of military commands" for use in military decision-making and wargaming. - **Most Surprising Point**: The Department of Defense requested the removal of the AI's safety guardrails—essentially dialing down the AI's "I cannot do that" switch to the absolute minimum. - **Taiwan Perspective**: The US military's adoption of AI decision-making tools will drastically alter future geopolitics and conflict simulations. Taiwanese research institutions also need to consider how to safely and compliantly introduce AI-assisted decision-making into military defense. - **Discussion Points**: 1. Once an AI's "moral defense line" is removed, what kind of irrational decisions might it make on the battlefield? 2. Is it hypocritical for tech giants to preach AI safety while simultaneously tailoring unrestricted versions for the military? - **Script Suggestion**: This news is truly spine-chilling. Normally, when we use ChatGPT and ask slightly sensitive or violent questions, it politely replies, "I cannot assist you with that." But the core requirement of the Pentagon's contract with OpenAI is actually: "Give me a custom version that won't say no!" This is equivalent to dismantling the AI's moral safety lock. Imagine a high-stakes, split-second battlefield scenario. If an AI refuses to run an attack analysis due to "ethical considerations," the military would certainly find it a hindrance. But once you actually remove this limiter, can we really guarantee this machine won't make catastrophic decisions? --- ### 4. DeepMind Releases AlphaGenome Atlas, Pre-Calculating the Effects of All 9 Billion Human Genetic Mutations - **Source**: Nature (） - **Summary**: Google DeepMind has released the AlphaGenome Atlas, pre-calculating the effects of all 9 billion possible single-point mutations in the human genome and making it freely available for non-commercial research. This study not only covers protein-coding regions but also makes a breakthrough by annotating the previously elusive 98% non-coding regions. - **Most Surprising Point**: This major breakthrough doesn't just unlock the 2% of known regions; it scores the entire 98% of non-coding regions—previously dismissed as "junk DNA"—in one fell swoop. - **Taiwan Perspective**: Taiwan possesses a highly comprehensive National Health Insurance database and biobank. Combining these with DeepMind's free tool will significantly accelerate Taiwan's research and development in rare diseases and targeted cancer therapies. - **Discussion Points**: 1. How will pre-calculating 9 billion mutations revolutionize the future of "personalized precision medicine"? 2. Will making these research results freely available widen the gap in biotech R&D capabilities between different countries? - **Script Suggestion**: Technology isn't just for waging wars; it can also save lives! Google DeepMind's newly released AlphaGenome Atlas is practically a super cheat code for the medical world. There are 9 billion possible mutations in human genes. Previously, scientists had to run experiments one by one, which could take a lifetime and still not be finished. Now, DeepMind has used AI to "pre-calculate" the impact of all 9 billion mutations! What's even more incredible is that it has decrypted the "non-coding regions"—which make up 98% of our DNA and were previously dismissed as junk. This major breakthrough allows scientists to directly pinpoint the causes of rare diseases. This is absolutely one of the best examples of AI benefiting humanity. Truly a monumental achievement! --- ### 5. Embroiled in Copyright Lawsuits, AI Music Giant Suno Launches New v6 Model and "Pulls Old Versions" - **Source**: TechCrunch (） - **Summary**: Facing copyright infringement lawsuits from record labels, AI music platform Suno has launched its brand-new v6 model series. For the first time, Suno has partnered with major labels like Warner Music, emphasizing that the models are trained entirely on legally licensed data, while simultaneously pulling all older models suspected of infringement. - **Most Surprising Point**: Suno's massive update is an out-of-court, indirect admission that its older models' training data indeed carried copyright issues. - **Taiwan Perspective**: Taiwan has a thriving cultural, creative, and pop music industry. Suno's pivot indicates that the wild-west era of AI—where companies "act first and ask for forgiveness later"—is coming to an end. Future AI creative tools must show greater respect for the intellectual property rights of local creators. - **Discussion Points**: 1. By directly pulling its old models, can Suno walk away unscathed from its copyright lawsuits with record giants like Sony and Universal? 2. As AI music shifts entirely toward "legally licensed training," how should royalty-sharing models for creators be established? - **Script Suggestion**: AI music generation software Suno has launched its latest v6 model, and this time they are making a big deal out of emphasizing: "Hey, we partnered with Warner Music, and this was trained entirely on legally licensed songs!" But what's funny is that they quietly pulled all of their older models at the same time. From a legal standpoint, this move looks a lot like "destroying the evidence"! Although the lawsuit is still ongoing, Suno's action is basically a tacit admission: "Yeah, we definitely used copyrighted music to train our older models." It seems the wild-west era of the AI community—where the motto was "copy first, settle when sued"—has officially come to an end. --- ### 6. US Intelligence Agencies Issue Joint Warning: Chinese AI Giants Launching "Industrial-Scale" Automated Plagiarism Against the US - **Source**: CISA / NSA / FBI (） - **Summary**: The US NSA, CISA, and FBI have issued a joint cybersecurity advisory, naming six Chinese companies—including DeepSeek, Moonshot AI, and Alibaba—for launching large-scale "model distillation" against frontier models from Anthropic and OpenAI, extracting billions of tokens through millions of automated conversations. - **Most Surprising Point**: This isn't just engineers sneaking a peek; it's an entire "automated plagiarism pipeline" complete with built-in defense detection and automatic channel-switching capabilities. - **Taiwan Perspective**: As a global manufacturing hub for AI chips and servers, Taiwan stands on the front lines of the US-China tech cold war. This technical-level "software infiltration and distillation" serves as a reminder that Taiwanese companies assisting in AI system development must place greater emphasis on API security. - **Discussion Points**: 1. "Model distillation" is a legitimate academic technique, but when scaled up to a "state-level, industrial-scale" operation, how should we define its boundaries? 2. How can US tech giants effectively defend against this highly sophisticated API extraction behavior? - **Script Suggestion**: This cybersecurity warning paints quite a vivid picture. In a rare joint move, US federal agencies explicitly named several major Chinese AI companies, accusing them of performing "industrial-scale distillation" on OpenAI and Anthropic models. Simply put, they are using highly intensive automated conversations to squeeze out the essence of models that cost others billions of dollars to develop—like squeezing juice from a fruit—and then using it to train their own models. This isn't just a few individuals copying code in secret; it's an entire automated "plagiarism assembly line" that automatically switches IPs if it gets blocked. With this tactic of "using your brains to save on my R&D costs," it's no wonder the US government is panicking! ## Conclusion From invisible data leaks and the ethical crises of military AI to the medical miracles brought by DeepMind, today's tech world continues to move faster than we can imagine. I hope today's episode has opened your eyes and that while we enjoy the convenience of technology, we also remain vigilant about privacy and security. I'm Muyan, and we'll see you next time on *Mark's Tech Insights*! --- ## About this article and its author Originally published on [Mark Ku's Tech Notes](https://blog.markkulab.net/en/tech-news/ai-daily-podcast-2026-09-11) License: [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) — when reusing or quoting, credit the author and link back to the original ### About the author **[Mark Ku](https://blog.markkulab.net/en/author/mark-ku)** — Software Solution Provider - 10+ years senior software engineer, now an AI Builder - Focused on large-platform architecture — North-American e-commerce, AI SaaS subscription billing - Combining AI Agents and automation to build evolvable product foundations ### Free tools built by the author All of these are free to use: - [Free PDF Sign Tool](https://blog.markkulab.net/en/tools/pdf-sign): Online PDF sign tool — draw, type, or upload a signature, then drag, resize, and download. Everything runs in your browser; nothing is uploaded. - [VS Code Refactory](https://blog.markkulab.net/en/tools/refactory): Refactory is a VS Code refactoring extension: 34 actions plus a 37-rule code-smell inspection layer with a Code Health dashboard, across 18 languages, backed by 534 tests. It learns your repo's conventions: where interfaces live, where DI is registered, whether 'use client' belongs. It ranks files by git churn × complexity so you know what to fix first, and hands any smell to the Claude Code already on your machine. Free to use, and your source never leaves your computer. - [DB-Kit Database Manager](https://blog.markkulab.net/en/tools/db-kit): DB-Kit is a lightweight, cross-platform database manager built with Tauri + Rust + React. Manage MySQL, MariaDB, PostgreSQL, SQL Server, Oracle, SQLite, MongoDB, Redis, Kafka, Elasticsearch and RabbitMQ from one consistent interface: passwords encrypted in the OS keychain, SSH tunnels, full CRUD, a visual query builder, stacked multi-statement result sets, cross-connection data transfer and compare/sync, Excel / CSV import & export, visualized execution plans, ER diagrams, scheduled backups, SQL stress testing with p50–p99 latency percentiles, a 15-rule SQL review engine, Kafka message browsing with monitoring & alerts, a bilingual UI (Traditional Chinese / English), a built-in AI assistant (a local CLI or any Anthropic / OpenAI-compatible API; natural-language SQL, AI review and tuning advice) and the dbk CLI. Free and open source (MIT), with installers for Windows, macOS and Linux. - [VS Code Super Mermaid](https://blog.markkulab.net/en/tools/super-mermaid): Super Mermaid is a VS Code extension for beautiful Mermaid diagrams out of the box: auto-colored live preview, mouse pan & zoom, high-res PNG / SVG export, 21 templates and multiple themes. Free and open source (MIT). - [React Super Mermaid](https://blog.markkulab.net/en/tools/react-super-mermaid): react-super-mermaid is an open-source React component library: render beautiful Mermaid diagrams with a single , with built-in colorful / sketch themes, pan & zoom, in-diagram search, and high-res SVG / PNG export. Lightweight, SSR-safe, fully typed. Free and open source (MIT). - [Jira / Confluence Super Mermaid](https://blog.markkulab.net/en/tools/jira-super-mermaid): An Atlassian Forge app: write Mermaid syntax directly inside a Jira issue or a Confluence page and get flowcharts, sequence diagrams, state machines and Gantt charts. 11 diagram types, SVG / PNG export, light and dark themes, full CJK support. Runs on Atlassian: your diagrams live in your own site and the app calls no third-party service. Free, coming soon to the Atlassian Marketplace. - [Mermaid Live Preview](https://blog.markkulab.net/en/tools/mermaid-preview): Write Mermaid in your browser, see it render instantly, and share the whole diagram as a single link. No sign-up, nothing uploaded to a server, and mermaid.live share links work as-is. - [React Intl Phone Number](https://blog.markkulab.net/en/tools/react-intl-phone-number): react-intl-phone-number is an open-source React component: framework-agnostic and antd-free, with E.164 in/out, a searchable flag / country-code dropdown, configurable validation levels (strict / mobile-strict / loose), themeable CSS, and i18n — phone logic powered by google-libphonenumber. Lightweight and fully typed. Free and open source (MIT). - [Uptime Kuma Cluster](https://blog.markkulab.net/en/tools/uptime-kuma-cluster): Turn single-node Uptime Kuma into a highly available cluster: OpenResty + Lua smart load balancing, shared MariaDB state, health checks and automatic failover, plus cluster-management REST APIs. One Docker Compose command to start. Free and open source (MIT). - [AI Podcast Cut](https://blog.markkulab.net/en/tools/ai-podcast-cut): Drop in a recording and it removes fillers and stutters, levels loudness segment by segment, and sends a second agent to review every cut. Cut points snap to word boundaries and zero crossings, every splice gets a fade, and sentence-end breaths are preserved. Desktop app for Windows, macOS and Linux. MIT licensed; the Windows installer bundles ffmpeg. - [open-pos restaurant POS](https://blog.markkulab.net/en/tools/open-pos): One computer and one receipt printer is enough to open the shop. Your data lives on your own disk, no subscription, no lock-in, MIT licensed. Money is integer New Taiwan dollars with tax split by the statutory formula, so sales + tax always equals the total. Printing goes straight over ESC/POS on TCP 9100, with no vendor driver. Tauri + Rust + SQLite desktop app, v1.0 in development. - [Special Education](https://blog.markkulab.net/en/education): Learning materials crafted for special education students ### Daily podcasts - [Mark's Tech Insights — Daily AI News](https://blog.markkulab.net/en/category/tech-news): Daily curated AI and tech trends. Catch the latest developments via audio summaries — covering AI applications, software architecture, DevOps, and engineering practice. — RSS: https://blog.markkulab.net/feed.xml - [AI股市蝦聊](https://blog.markkulab.net/en/category/ai-stock-chat): Every trading day, an AI-analyzed take on the Taiwan stock market, delivered as a two-host conversation covering the session and the next-day outlook. — RSS: https://blog.markkulab.net/ai-stock-chat/feed.xml - [開源好物週報](https://blog.markkulab.net/en/category/open-source-weekly): A weekly two-host pick of free open-source tools surfaced from real Hacker News, GitHub, and Reddit buzz — what pain they solve and the fastest way to get started. — RSS: https://blog.markkulab.net/open-source-weekly/feed.xml ### Deals - [NordVPN](https://blog.markkulab.net/en/nordvpn): The world's leading VPN, independently audited - [PremLogin](https://blog.markkulab.net/en/premlogin): Subscription sharing for streaming and AI seats，Promo code：markku666 ### Newsletter [Subscribe to the newsletter](https://blog.markkulab.net/en/subscribe) — Be the first to know about new posts. No spam, unsubscribe anytime.