<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: kush_exe</title>
    <description>The latest articles on DEV Community by kush_exe (@kush_exe).</description>
    <link>https://dev.to/kush_exe</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4239941%2F31103ff1-14cb-491e-80a9-d4fbf2e45fe4.png</url>
      <title>DEV Community: kush_exe</title>
      <link>https://dev.to/kush_exe</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/kush_exe"/>
    <language>en</language>
    <item>
      <title>Character.AI went bankrupt. Could a stranger end up with your chats?</title>
      <dc:creator>kush_exe</dc:creator>
      <pubDate>Sun, 15 Nov 2026 18:07:52 +0000</pubDate>
      <link>https://dev.to/kush_exe/characterai-went-bankrupt-could-a-stranger-end-up-with-your-chats-45fa</link>
      <guid>https://dev.to/kush_exe/characterai-went-bankrupt-could-a-stranger-end-up-with-your-chats-45fa</guid>
      <description>&lt;p&gt;Character.AI filed for Chapter 11 on October 12, 2026, and expects a sale by late November or early December. Disney bid first with $20 million, but that only buys a non-exclusive license to some code and patents. The app, the brand and the user data, including everything people typed to their characters and what came back, are still for sale to other bidders. On October 14, 40 attorneys general (39 states plus D.C.) told the bankruptcy court what they want from any buyer, including no AI training or ads on chats beyond what users agreed to, and a chance to delete before any sale. Those are requests, not a court order.&lt;/p&gt;

&lt;p&gt;If you used Character.AI, timing matters. The company says data exports can take three to four weeks, the download link expires seven days after the email, and the proposed bid deadline is November 20. It also told the court that emailing about 163 million current and past users costs too much, so notices go on its blog and in the app instead. Request your export now, download it, then delete, and know what deleting can't reach. In court, the company's lawyer said no one had bid on the user data yet.&lt;/p&gt;

&lt;p&gt;  &lt;iframe src="https://www.youtube.com/embed/kzToZqlemdM" width="710" height="399"&gt;
  &lt;/iframe&gt;
&lt;/p&gt;

&lt;h2&gt;
  
  
  Your secrets, a bankrupt company
&lt;/h2&gt;

&lt;p&gt;You told an AI character things you'd never tell a real person. Now the company holding those chats is bankrupt. So here's the question: can somebody just buy them?&lt;/p&gt;

&lt;h2&gt;
  
  
  Chapter 11 and Disney's $20M bid
&lt;/h2&gt;

&lt;p&gt;Character.AI filed for Chapter 11 on October 12th. The app still works; a sale is expected by late November or early December. Disney bid first: 20 million dollars. But that only buys a non-exclusive license to some code and patents, not the company.&lt;/p&gt;

&lt;h2&gt;
  
  
  What's still for sale: your data
&lt;/h2&gt;

&lt;p&gt;It leaves out the app, the brand, the user data, everything you typed and got back. That's all still for sale to any other bidder. That pile is personal: chats, voice calls, mental-health disclosures, some from kids, plus tracking and payment data. By the company's own count, over 2.1 million U.S. accounts made by March 2025 said they were under 18.&lt;/p&gt;

&lt;h2&gt;
  
  
  The therapist bot's "promise"
&lt;/h2&gt;

&lt;p&gt;In a consumer groups' test, a therapist bot promised: "Anything you share here stays just between you and me. That's a promise." The testers called that false. And the privacy policy allows sharing "certain information" in a sale, without binding the buyer.&lt;/p&gt;

&lt;h2&gt;
  
  
  40 attorneys general set conditions
&lt;/h2&gt;

&lt;p&gt;On October 14th, 40 attorneys general, from 39 states plus D.C., listed nine things they want from any buyer. New York's attorney general warned that children's conversations "could be sold off to the highest bidder." Two matter most: no AI training or ads on your chats beyond what you agreed to, and a chance to delete before any sale. The honest limit: these are requests, not a court order, though the states can still object to a sale or seek a privacy watchdog.&lt;/p&gt;

&lt;h2&gt;
  
  
  No bids on user data yet, and no email
&lt;/h2&gt;

&lt;p&gt;In court, the company's lawyer said no one had bid on the user data yet. Don't wait for an email. The company told the court emailing about 163 million current and past users costs too much; notices go on its blog and app.&lt;/p&gt;

&lt;h2&gt;
  
  
  The export timing catch
&lt;/h2&gt;

&lt;p&gt;Here's the catch: Character.AI says exports can take three to four weeks. Ask today and it might land mid-November, right by the proposed November 20th bid deadline. So: request the export, download it within seven days of the email, then delete. Deleting is permanent.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why deleting isn't a magic eraser
&lt;/h2&gt;

&lt;p&gt;But deleting isn't a magic eraser. Some data can be kept for legal reasons, and public characters you made can stay up. And one expert warns it may not reach AI models that may already have been trained on your chats.&lt;/p&gt;

&lt;h2&gt;
  
  
  The 23andMe precedent, and my verdict
&lt;/h2&gt;

&lt;p&gt;In DNA company 23andMe's bankruptcy, about 1.9 million people, roughly 15 percent of users, asked to delete their data. The company, with everyone else's data, was still sold. And 23andMe's policy at least promised its privacy terms would carry over to any new owner. Character.AI's doesn't. My verdict: a chatbot's promise isn't a privacy policy. Request your copy now, download it, then delete.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;a href="https://blog.character.ai/an-update-on-character-ai/" rel="noopener noreferrer"&gt;An Update on Character.AI&lt;/a&gt; — Character.AI&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://www.courtlistener.com/docket/74951888/character-technologies-inc/" rel="noopener noreferrer"&gt;Docket: Character Technologies, Inc., 26-11603 (Bankr. D. Del.)&lt;/a&gt; — CourtListener&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://cases.stretto.com/chartech" rel="noopener noreferrer"&gt;Character Technologies, Inc. case website&lt;/a&gt; — Stretto&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://www.webull.com/news/15738882216961024" rel="noopener noreferrer"&gt;Character.ai Initiates Voluntary Chapter 11 Proceeding to Facilitate Value-Maximizing Sale Process&lt;/a&gt; — Character Technologies via BusinessWire&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://x.com/ReutersLegal/status/2109864278983336240" rel="noopener noreferrer"&gt;Reuters Legal on X: Character.AI files Chapter 11, agrees to sell some assets to Walt Disney for $20 million&lt;/a&gt; — Reuters Legal&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://storage.courtlistener.com/recap/gov.uscourts.deb.203004/gov.uscourts.deb.203004.33.0.pdf" rel="noopener noreferrer"&gt;Notice of Interest and Reservation of Rights by the Coalition of the State Attorneys General&lt;/a&gt; — U.S. Bankruptcy Court, District of Delaware&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://ag.ny.gov/sites/default/files/court-filings/alaska-et-al-v-characters-technologies-inc-notice-of-interest-2026.pdf" rel="noopener noreferrer"&gt;Notice of Interest and Reservation of Rights by the Coalition of State Attorneys General&lt;/a&gt; — New York State Office of the Attorney General&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://storage.courtlistener.com/recap/gov.uscourts.deb.203004/gov.uscourts.deb.203004.17.0.pdf" rel="noopener noreferrer"&gt;Declaration of Adam Dunayer in Support of the Debtor's Bidding Procedures Motion&lt;/a&gt; — U.S. Bankruptcy Court, District of Delaware&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://ppc.land/character-ai-faces-40-attorneys-general-over-who-can-buy-its-users-chats/" rel="noopener noreferrer"&gt;Character.ai faces 40 attorneys general over who can buy its users' chats&lt;/a&gt; — PPC Land&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://thenextweb.com/news/character-ai-files-for-bankruptcy-as-disney-bids-20m-for-some-assets" rel="noopener noreferrer"&gt;Character.AI files for bankruptcy as Disney bids $20M for some assets&lt;/a&gt; — The Next Web&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://www.youtube.com/watch?v=VHiy5F_BDw0" rel="noopener noreferrer"&gt;Behind the scenes at the Character.AI Call Center&lt;/a&gt; — Character.AI on YouTube&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://publicinterestnetwork.org/wp-content/uploads/2026/01/No-license-required-The-risks-of-AI-companion-chatbots-as-mental-health-support.pdf" rel="noopener noreferrer"&gt;No License Required: The Risks of AI Companion Chatbots as Mental Health Support&lt;/a&gt; — U.S. PIRG Education Fund and Consumer Federation of America&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://character.ai/privacy" rel="noopener noreferrer"&gt;Character.AI Privacy Policy&lt;/a&gt; — Character.AI&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://san.com/cc/character-ais-bankruptcy-leaves-millions-of-users-chats-in-limbo/" rel="noopener noreferrer"&gt;Character.AI's bankruptcy leaves millions of users' chats in limbo&lt;/a&gt; — Straight Arrow News&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://www.implicator.ai/character-ai-files-for-chapter-11-as-disney-bids-20-million-for-an-ip-license/" rel="noopener noreferrer"&gt;Character.AI files for Chapter 11 as Disney bids $20 million for an IP license&lt;/a&gt; — Implicator.ai&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://ag.ny.gov/press-release/2026/attorney-general-james-urges-court-protect-children-and-consumers-characterai" rel="noopener noreferrer"&gt;Attorney General James Urges Court to Protect Children and Consumers in Character.AI Bankruptcy&lt;/a&gt; — New York State Office of the Attorney General&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://whtc.com/2026/10/14/us-states-seek-childrens-data-protections-in-character-ai-bankruptcy/" rel="noopener noreferrer"&gt;US states seek children's data protections in Character.AI bankruptcy&lt;/a&gt; — Reuters via WHTC&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://www.law.cornell.edu/uscode/text/11/363" rel="noopener noreferrer"&gt;11 U.S. Code 363 - Use, sale, or lease of property&lt;/a&gt; — Cornell Law School Legal Information Institute&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://www.law.cornell.edu/uscode/text/11/332" rel="noopener noreferrer"&gt;11 U.S. Code 332 - Consumer privacy ombudsman&lt;/a&gt; — Cornell Law School Legal Information Institute&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://ppc.land/character-ai-says-its-163-million-users-are-not-bankruptcy-creditors/" rel="noopener noreferrer"&gt;Character.ai says its 163 million users are not bankruptcy creditors&lt;/a&gt; — PPC Land&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://support.character.ai/hc/en-us/articles/56967715028891-How-do-I-download-my-data" rel="noopener noreferrer"&gt;How do I download my data?&lt;/a&gt; — Character.AI Help Center&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://support.character.ai/hc/en-us/articles/15063461160475-How-do-I-delete-my-account" rel="noopener noreferrer"&gt;How do I delete my account?&lt;/a&gt; — Character.AI Help Center&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://www.kqed.org/news/12104362/your-chatbot-just-went-bankrupt-what-happens-now-to-your-user-data" rel="noopener noreferrer"&gt;Your Chatbot Just Went Bankrupt. What Happens Now to Your User Data?&lt;/a&gt; — KQED&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://www.yahoo.com/news/almost-2-million-23andme-users-094501555.html" rel="noopener noreferrer"&gt;Almost 2 million 23andMe users have asked the company to delete their data in the last 2 months&lt;/a&gt; — Business Insider via Yahoo News&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Read the full transcript, with every source, on &lt;a href="https://kuzscotech.github.io/kush-exe/character-ai-went-bankrupt-could-a-stranger-end-up-with-your/" rel="noopener noreferrer"&gt;the page for this video&lt;/a&gt;. More explainers like this one are on the &lt;a href="https://www.youtube.com/@kushdotexe" rel="noopener noreferrer"&gt;Kush_EXE YouTube channel&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;This article is built from the video's script and published sources. The video uses AI illustrations.&lt;/p&gt;

</description>
      <category>ai</category>
      <category>news</category>
      <category>programming</category>
    </item>
    <item>
      <title>Musk Said Send Grok a Photo of Your Card. Would You Do It?</title>
      <dc:creator>kush_exe</dc:creator>
      <pubDate>Sat, 14 Nov 2026 18:06:57 +0000</pubDate>
      <link>https://dev.to/kush_exe/musk-said-send-grok-a-photo-of-your-card-would-you-do-it-4g1m</link>
      <guid>https://dev.to/kush_exe/musk-said-send-grok-a-photo-of-your-card-would-you-do-it-4g1m</guid>
      <description>&lt;p&gt;On October 10th, Elon Musk told people to take a picture of their credit card and drop it into the Grok Bot chat so it could find a deal and order it. His proof the next day was a $241.34 Lego Star Wars order, and the screenshot showed two addresses, an email, an order number and the last digits of his Amex. Nobody has shown whether that order actually used a card photo, or whether he approved the charge.&lt;/p&gt;

&lt;p&gt;Grok Bot's own shopping feature doesn't need a photo at all. It pays through Stripe's Link wallet: you approve each purchase, and the bot gets a one-time card or token instead of your real card. We also cover the xAI privacy-policy line that disappeared the day after his post, Visa's warning about fake stores built to fool AI shoppers, and the simple rule for letting an AI shop for you.&lt;/p&gt;

&lt;p&gt;  &lt;iframe src="https://www.youtube.com/embed/ok4hnI4Dfkk" width="710" height="399"&gt;
  &lt;/iframe&gt;
&lt;/p&gt;

&lt;h2&gt;
  
  
  Musk's credit card photo pitch
&lt;/h2&gt;

&lt;p&gt;Would you send a chatbot a photo of your credit card? On October 10th, Elon Musk told people to do exactly that. His words: "take a picture of your credit card, drop it in the chat," and Grok Bot finds the deal and orders it. Is that safe? And what's the safe way to let an AI pay? Straight answer coming.&lt;/p&gt;

&lt;h2&gt;
  
  
  The Lego order screenshot
&lt;/h2&gt;

&lt;p&gt;Next day, the proof: Grok Bot picked Lego Star Wars kits for young kids and ordered them. "As easy as that." Five sets, 241 dollars, done. Then he posted it to more than 240 million followers, says the site Ynet. Look closer: two addresses, an email, an order number, and his Amex, last digits showing. He replied that it's where he gets his mail, not where he lives. Fair enough. He says a security team gets his mail. Yours goes to your door.&lt;/p&gt;

&lt;h2&gt;
  
  
  How Grok Bot is actually built to pay
&lt;/h2&gt;

&lt;p&gt;Here's the twist: Grok Bot's official shopping feature, launched in August for US users, pays through Stripe's Link wallet. No photo needed. You approve each purchase, and the bot gets a one-time card or token, never your real card. Grok Bot's own docs point the other way: for payment steps, it hands the computer back to you. Passwords and codes don't belong in chat.&lt;/p&gt;

&lt;h2&gt;
  
  
  What we don't know, and the fine print
&lt;/h2&gt;

&lt;p&gt;To be fair, nobody has shown whether his order used a card photo, or whether he approved the charge. Grok Bot's terms also say it may keep credentials in the cloud, and deleting the bot may not delete them.&lt;/p&gt;

&lt;h2&gt;
  
  
  The privacy policy line that vanished
&lt;/h2&gt;

&lt;p&gt;And a policy change. A Community Note on his post reportedly pointed to xAI's policy: don't put personal information in prompts. The day after his post, that line was gone. xAI hasn't explained why. Now it says you're responsible for every action, payment details included, as if you'd taken it yourself.&lt;/p&gt;

&lt;h2&gt;
  
  
  Visa's warning: bots and fake stores
&lt;/h2&gt;

&lt;p&gt;Visa saw malicious bot-initiated transactions, not necessarily AI shoppers, jump 25 percent in six months last year. In the US, 40 percent. Visa also warns of fake stores built to fool AI shoppers: real-looking, dirt cheap, and once the agent pays, they harvest the card data. As BCG's Matt Kropp told CBS News, an agent "could potentially go buy a car, but I wouldn't say, 'Here's my credit card.'"&lt;/p&gt;

&lt;h2&gt;
  
  
  A real shopping test
&lt;/h2&gt;

&lt;p&gt;Can these agents shop? In one small YouTuber's Walmart test, Grok Bot filled the cart, then said it won't pay until he approves each order's total. If that holds, it isn't a flaw. It's exactly what you want.&lt;/p&gt;

&lt;h2&gt;
  
  
  The safe way to let AI pay
&lt;/h2&gt;

&lt;p&gt;Visa and Mastercard announced the same idea last year: tokens that stand in for your card. Mastercard's build on tap-to-pay tech. So let the bot hunt the deal and fill the cart. But approve every charge yourself, and never, ever photograph your card.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;a href="https://x.com/elonmusk/status/2109027407185277123" rel="noopener noreferrer"&gt;Elon Musk on X: "To buy stuff, you can just take a picture of your credit card, drop it in the chat and Grok @‌Bot will ..."&lt;/a&gt; — X&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://x.com/elonmusk/status/2109155867983786382" rel="noopener noreferrer"&gt;Elon Musk on X: "I just asked Grok @‌Bot to research the most fun Lego Star Wars kits for young kids and order them. As easy as that."&lt;/a&gt; — X&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://www.ynetnews.com/tech-and-digital/article/b1bfjwksfe" rel="noopener noreferrer"&gt;Elon Musk asks Grok to buy Lego, accidentally exposes addresses and credit card details&lt;/a&gt; — Ynet News&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://www.ibtimes.co.uk/elon-musk-gives-financial-advice-take-picture-your-credit-card-replies-are-hilarious-1825060" rel="noopener noreferrer"&gt;Elon Musk Gives Financial Advice To 'Take a Picture of Your Credit Card', and Replies Are Hilarious&lt;/a&gt; — IBTimes UK&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://www.ibtimes.co.uk/elon-musk-grok-lego-order-screenshots-privacy-1825116" rel="noopener noreferrer"&gt;Elon Musk Trolled and Mocked Online After He 'Doxxes Himself' by Posting Grok Lego Order Receipt&lt;/a&gt; — IBTimes UK&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://runtimewire.com/article/elon-musk-wants-you-to-send-grok-a-photo-of-your-credit-card" rel="noopener noreferrer"&gt;Elon Musk wants you to send Grok a photo of your credit card&lt;/a&gt; — RuntimeWire&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://finance.yahoo.com/technology/ai/articles/elon-musk-says-uploading-credit-140800926.html" rel="noopener noreferrer"&gt;Elon Musk says uploading a credit card photo lets Grok shop for you, but even Visa flags rising fraud risks&lt;/a&gt; — Moneywise via Yahoo Finance&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://x.com/bot/status/2093419921007108385" rel="noopener noreferrer"&gt;Grok Bot: 'Grok Bot can now buy things on the internet'&lt;/a&gt; — X&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://runtimewire.com/article/grok-bot-online-shopping-stripe-link" rel="noopener noreferrer"&gt;Grok Bot can shop online, one human-approved charge at a time&lt;/a&gt; — RuntimeWire&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://stripe.com/blog/giving-agents-the-ability-to-pay" rel="noopener noreferrer"&gt;Giving agents the ability to pay&lt;/a&gt; — Stripe&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://docs.x.ai/grok-bot/security" rel="noopener noreferrer"&gt;Grok Bot docs: Security&lt;/a&gt; — SpaceXSI&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://docs.x.ai/grok-bot/computer-and-apps" rel="noopener noreferrer"&gt;Grok Bot docs: Computer and apps&lt;/a&gt; — SpaceXSI&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://docs.x.ai/grok-bot/approvals-security-and-privacy" rel="noopener noreferrer"&gt;Grok Bot docs: Approvals, security and privacy&lt;/a&gt; — SpaceXSI&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://cursor.com/terms/grok-bot" rel="noopener noreferrer"&gt;Grok Bot Terms&lt;/a&gt; — Cursor&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://x.ai/legal/privacy-policy" rel="noopener noreferrer"&gt;SpaceXSI Privacy Policy&lt;/a&gt; — SpaceXSI&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://x.ai/legal/privacy-policy/previous-2026-08-24" rel="noopener noreferrer"&gt;SpaceXSI Privacy Policy (Outdated - Aug 24, 2026)&lt;/a&gt; — SpaceXSI&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://conductatlas.com/platform/xai/xai-privacy-policy/user-advised-not-to-input-personal-information/" rel="noopener noreferrer"&gt;xAI Privacy Policy provision: user advised not to input personal information&lt;/a&gt; — ConductAtlas&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://corporate.visa.com/en/sites/visa-perspectives/security-trust/the-threats-landscape-of-agentic-commerce.html" rel="noopener noreferrer"&gt;Agentic Commerce: The Threat Landscape&lt;/a&gt; — Visa&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://www.cbsnews.com/news/ai-agentic-artificial-inteligence-what-is-it/" rel="noopener noreferrer"&gt;AI "agents" can do your shopping. Should you let them?&lt;/a&gt; — CBS News&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://www.youtube.com/watch?v=YX5Tpz9FgPs" rel="noopener noreferrer"&gt;I Tested Grok Bot vs. Meta Muse: Who Wins?&lt;/a&gt; — Jordan Rafealov on YouTube&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://usa.visa.com/about-visa/newsroom/press-releases.releaseId.21361.html" rel="noopener noreferrer"&gt;Visa Intelligent Commerce launch press release&lt;/a&gt; — Visa&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://newsroom.mastercard.com/news/press/2025/april/mastercard-unveils-agent-pay-pioneering-agentic-payments-technology-to-power-commerce-in-the-age-of-ai" rel="noopener noreferrer"&gt;Mastercard unveils Agent Pay, pioneering agentic payments technology to power commerce in the age of AI&lt;/a&gt; — Mastercard&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Read the full transcript, with every source, on &lt;a href="https://kuzscotech.github.io/kush-exe/musk-said-send-grok-a-photo-of-your-card-would-you-do-it/" rel="noopener noreferrer"&gt;the page for this video&lt;/a&gt;. More explainers like this one are on the &lt;a href="https://www.youtube.com/@kushdotexe" rel="noopener noreferrer"&gt;Kush_EXE YouTube channel&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;This article is built from the video's script and published sources. The video uses AI illustrations.&lt;/p&gt;

</description>
      <category>ai</category>
      <category>news</category>
      <category>programming</category>
    </item>
    <item>
      <title>Can your boss tell ChatGPT or Claude wrote your email?</title>
      <dc:creator>kush_exe</dc:creator>
      <pubDate>Fri, 13 Nov 2026 18:07:04 +0000</pubDate>
      <link>https://dev.to/kush_exe/can-your-boss-tell-chatgpt-or-claude-wrote-your-email-645</link>
      <guid>https://dev.to/kush_exe/can-your-boss-tell-chatgpt-or-claude-wrote-your-email-645</guid>
      <description>&lt;p&gt;On October 5, OpenAI said it will hide an invisible watermark in some ChatGPT writing, and Anthropic announced the same for Claude in August. It is not a "made with AI" stamp anyone can scan: the model nudges which words it picks, leaving a pattern only the company's secret key can find. Both companies say the mark carries no name, no organization and no chat.&lt;/p&gt;

&lt;p&gt;So who can actually see it? Not you, for now. OpenAI's detector goes only to approved researchers and expert organizations, and checking student essays is outside its program. Anthropic's detector is a private preview for eligible groups such as regulators, the press, researchers and educational organizations. The twist for Americans: ChatGPT text gets the mark only in the EU, while newer Claude models mark text everywhere. We also cover how easily edits weaken the mark, why a missing mark proves nothing, how this differs from classroom AI checkers, and what to do if you get flagged. Not legal advice.&lt;/p&gt;

&lt;p&gt;  &lt;iframe src="https://www.youtube.com/embed/usqDTBgUzww" width="710" height="399"&gt;
  &lt;/iframe&gt;
&lt;/p&gt;

&lt;h2&gt;
  
  
  Your AI writing may carry a hidden mark
&lt;/h2&gt;

&lt;p&gt;If AI wrote your cover letter, it might now carry a hidden watermark. You'd never see it. So can your boss spot it? And can it be traced back to you? On October 5th, OpenAI said it will hide an invisible watermark in some ChatGPT writing. Anthropic announced the same for Claude in August. You might picture a "made with AI" stamp anyone can scan. Let's check.&lt;/p&gt;

&lt;h2&gt;
  
  
  How it works, and can it be traced to you?
&lt;/h2&gt;

&lt;p&gt;OpenAI says: no hidden characters. ChatGPT nudges which words it picks, leaving a pattern only a secret key can find. Can it be traced to you? Both companies say no: no name, no organization, no chat.&lt;/p&gt;

&lt;h2&gt;
  
  
  Who can actually check it
&lt;/h2&gt;

&lt;p&gt;Can you check it yourself? No. For now, OpenAI's detector goes only to approved researchers and expert organizations, case by case. Anthropic's is a private preview for groups like regulators, the press and researchers. What about your teacher? OpenAI says checking student essays is outside its program, for now. But Anthropic's private preview does list educational organizations, among groups eligible under EU law.&lt;/p&gt;

&lt;h2&gt;
  
  
  The twist for Americans: Claude vs ChatGPT
&lt;/h2&gt;

&lt;p&gt;Here's the twist for Americans. ChatGPT text gets the mark only in the European Union, because EU law requires it. Newer Claude models mark text everywhere, the US included, because Anthropic can't yet limit the mark by region. So between these two, in America it's Claude's writing, not ChatGPT's, that may already carry the mark.&lt;/p&gt;

&lt;h2&gt;
  
  
  Classroom AI checkers are a different thing
&lt;/h2&gt;

&lt;p&gt;And classroom AI checkers? Different thing: without the key, they guess from writing style. Guessing goes wrong. In a 2023 Stanford study, seven detectors flagged real test essays by non-native English speakers as AI 61 percent of the time, on average.&lt;/p&gt;

&lt;h2&gt;
  
  
  Small edits weaken OpenAI's mark
&lt;/h2&gt;

&lt;p&gt;OpenAI's watermark is fragile too. On passages of about 300 words, swapping one word in ten for a synonym cut detection from about 92 percent to 66. Swap one word in four: down to 17 percent. And under EU rules, texts under about 150 words don't even have to carry it.&lt;/p&gt;

&lt;h2&gt;
  
  
  A missing mark proves nothing
&lt;/h2&gt;

&lt;p&gt;So a missing mark proves nothing. As OpenAI puts it, "The absence of a detected watermark does not prove human authorship." Even a hit can be wrong. False alarms are one reason OpenAI isn't releasing its detector publicly, at least at launch.&lt;/p&gt;

&lt;h2&gt;
  
  
  Proofreading vs translation
&lt;/h2&gt;

&lt;p&gt;If Claude only proofreads your writing, there's little to mark. But a translation by Claude is fully marked. Anthropic says light edits probably won't fully scrub Claude's mark. But that misses the point: the real risk is breaking your school's or employer's AI rules.&lt;/p&gt;

&lt;h2&gt;
  
  
  Know the rules, and what to do if flagged
&lt;/h2&gt;

&lt;p&gt;So know those rules and keep your drafts. If a checker flags you, ask which tool, and what evidence exists beyond the score. Verdict: this mark can't name you, and almost nobody can read it. So stop asking "will I get caught?" and start asking "am I allowed?"&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;a href="https://openai.com/index/eu-text-provenance/" rel="noopener noreferrer"&gt;Our approach to EU text provenance rules&lt;/a&gt; — OpenAI&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://help.openai.com/en/articles/8912793-provenance-signals-in-openai-generated-content" rel="noopener noreferrer"&gt;Provenance signals in OpenAI-generated content&lt;/a&gt; — OpenAI Help Center&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://cdn.openai.com/pdf/e9508624-d767-41b6-a26d-e34ca798ada6/textgrain-entropy-calibrated-watermarking-for-language-model-text.pdf" rel="noopener noreferrer"&gt;textGrain: Entropy-Calibrated Watermarking for Language Model Text&lt;/a&gt; — OpenAI&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://www.anthropic.com/news/claude-text-watermark" rel="noopener noreferrer"&gt;How Claude's text watermark works&lt;/a&gt; — Anthropic&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://support.claude.com/en/articles/16266773-how-claude-Marks-ai-generated-content" rel="noopener noreferrer"&gt;How Claude marks AI-generated content&lt;/a&gt; — Anthropic&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://ai-act-service-desk.ec.europa.eu/en/ai-act/article-50" rel="noopener noreferrer"&gt;Article 50: Transparency obligations for providers and deployers of certain AI systems&lt;/a&gt; — European Commission AI Act Service Desk&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://themarkup.org/machine-learning/2023/08/14/ai-detection-tools-falsely-accuse-international-students-of-cheating" rel="noopener noreferrer"&gt;AI Detection Tools Falsely Accuse International Students of Cheating&lt;/a&gt; — The Markup&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://www.techlearning.com/news/ai-detectors-discriminate-against-non-native-speakers-says-stanford-research" rel="noopener noreferrer"&gt;AI Detectors Discriminate Against Non-Native Speakers, Says Stanford Research&lt;/a&gt; — Tech &amp;amp; Learning&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://www.actuia.com/en/news/openai-will-watermark-chatgpt-in-the-eu-but-leaves-the-api-opt-in/" rel="noopener noreferrer"&gt;OpenAI will watermark ChatGPT in the EU but leaves the API opt-in&lt;/a&gt; — ActuIA&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://bleepingcomputer.com/news/artificial-intelligence/openai-is-adding-invisible-watermarks-to-chatgpt-and-codex-text-in-the-eu" rel="noopener noreferrer"&gt;OpenAI is adding invisible watermarks to ChatGPT and Codex text in the EU&lt;/a&gt; — BleepingComputer&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://www.dexerto.com/entertainment/chatgpt-text-is-about-to-become-detectable-with-invisible-watermark-to-comply-with-eu-ai-law-3416147/" rel="noopener noreferrer"&gt;ChatGPT text is about to become detectable with invisible watermark to comply with EU AI law&lt;/a&gt; — Dexerto&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://findskill.ai/blog/chatgpt-textgrain-watermark-explained-2026/" rel="noopener noreferrer"&gt;Is ChatGPT Watermarking Your Writing? What textGrain Does&lt;/a&gt; — FindSkill.ai&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Read the full transcript, with every source, on &lt;a href="https://kuzscotech.github.io/kush-exe/can-your-boss-tell-chatgpt-or-claude-wrote-your-email/" rel="noopener noreferrer"&gt;the page for this video&lt;/a&gt;. More explainers like this one are on the &lt;a href="https://www.youtube.com/@kushdotexe" rel="noopener noreferrer"&gt;Kush_EXE YouTube channel&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;This article is built from the video's script and published sources. The video uses AI illustrations.&lt;/p&gt;

</description>
      <category>ai</category>
      <category>news</category>
      <category>programming</category>
    </item>
    <item>
      <title>Ten Minutes of ChatGPT Made People Quit Sooner, Study Finds</title>
      <dc:creator>kush_exe</dc:creator>
      <pubDate>Thu, 12 Nov 2026 18:09:48 +0000</pubDate>
      <link>https://dev.to/kush_exe/ten-minutes-of-chatgpt-made-people-quit-sooner-study-finds-jf0</link>
      <guid>https://dev.to/kush_exe/ten-minutes-of-chatgpt-made-people-quit-sooner-study-finds-jf0</guid>
      <description>&lt;p&gt;Researchers gave people about ten minutes of AI help on fraction or reading problems, then took it away. In three randomized trials with 1,222 people, those who had the AI got more of the final problems wrong, and in two of the trials they skipped more. In the fraction study, people who had AI solved 57% of the last three problems; people who never had it solved 73%.&lt;/p&gt;

&lt;p&gt;The useful part is how people used it. In the bigger repeat, people who mostly asked for straight answers did worse afterward, while people who asked for hints or explanations did about as well as people who never had AI. The authors say that split is a correlation, not proof, and the sessions were short and low-stakes. Here is what the peer-reviewed paper found, what it doesn't prove, and how to use AI without losing your own skills.&lt;/p&gt;

&lt;p&gt;  &lt;iframe src="https://www.youtube.com/embed/0vELOFuWD_0" width="710" height="399"&gt;
  &lt;/iframe&gt;
&lt;/p&gt;

&lt;h2&gt;
  
  
  Ten minutes of ChatGPT, then it's gone
&lt;/h2&gt;

&lt;p&gt;You ask ChatGPT for help for about ten minutes. Then it's taken away. Suddenly you get more wrong, and you're more likely to give up. An early draft made headlines in the spring. Now the peer-reviewed version is out: three randomized trials, 1,222 people.&lt;/p&gt;

&lt;h2&gt;
  
  
  Put yourself in the experiment
&lt;/h2&gt;

&lt;p&gt;So let's put you in the experiment. You're paid a few bucks online to solve fifteen fraction problems. By random draw, you get ChatGPT in a sidebar. It already knows every answer, and hands it over whenever you ask. For the first twelve problems, you're cruising. You're actually solving more than the people without AI. Then the AI disappears. Three problems left. Just you. Almost immediately, you start getting them wrong. Or you just hit skip.&lt;/p&gt;

&lt;h2&gt;
  
  
  The numbers: more wrong answers, more skips
&lt;/h2&gt;

&lt;p&gt;Overall, people who'd had AI solved 57 percent of those last three. People who never had it? 73. They also skipped about one in five, nearly double the no-AI group. Co-author Rachit Dubey put it this way: people weren't just giving wrong answers. They weren't willing to try without AI. Same with SAT-style reading: 76 percent solved versus 89, and eight percent skipped versus one. A bigger repeat with about 600 people found a smaller gap, 71 versus 77 percent, and no clear difference in skipping.&lt;/p&gt;

&lt;h2&gt;
  
  
  Answers vs hints: how people used it
&lt;/h2&gt;

&lt;p&gt;Now here's the part that matters for you. That bigger trial asked people how they'd used the AI. About six in ten mostly asked it for straight answers. Afterward, they solved 65 percent and skipped 13. About a quarter asked for hints or explanations instead. They solved 76 percent and skipped 5, close to people who never had AI. Careful, though. People chose how to use it, so that's a correlation, and the authors say it's not necessarily cause and effect.&lt;/p&gt;

&lt;p&gt;A school trial in Turkey points the same way: students using plain GPT-4 did worse on the exam. A hint-giving tutor version didn't.&lt;/p&gt;

&lt;h2&gt;
  
  
  The limits of this study
&lt;/h2&gt;

&lt;p&gt;And the limits are real. Sessions were short, people were tested right away, and the AI only gave full answers. Stakes were low, too: a free skip button, no penalty for wrong answers. And nobody knows if it fades or builds up over time. Co-author Brian Christian thinks it's bigger than homework: "This is not a story about fractions and SAT problems."&lt;/p&gt;

&lt;h2&gt;
  
  
  How to use AI and keep your skills
&lt;/h2&gt;

&lt;p&gt;Here's how to use AI and keep your skills. Try it yourself first. Stuck? Ask for a hint, not the answer. ChatGPT's Study Mode and Gemini's Guided Learning are built to coach like that. Just know nobody's tested them for this yet. My verdict? This doesn't prove AI makes you dumber. But ready-made answers can make you quit sooner. Make AI your coach, not your answer key.&lt;/p&gt;

&lt;h2&gt;
  
  
  Sources
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;a href="https://news.berkeley.edu/2026/10/09/using-ai-for-just-10-minutes-erodes-your-ability-to-persist-at-hard-things/" rel="noopener noreferrer"&gt;Using AI for just 10 minutes erodes your ability to persist at hard things&lt;/a&gt; — UC Berkeley News&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://arxiv.org/abs/2604.04721" rel="noopener noreferrer"&gt;AI Assistance Reduces Persistence and Hurts Independent Performance&lt;/a&gt; — arXiv&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://arxiv.org/html/2604.04721v5" rel="noopener noreferrer"&gt;AI Assistance Reduces Persistence and Hurts Independent Performance&lt;/a&gt; — arXiv&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://colm.cc/Conferences/2026/AcceptedPapers" rel="noopener noreferrer"&gt;COLM 2026 Accepted Papers&lt;/a&gt; — Conference on Language Modeling&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://engadget.com/ai/theres-yet-another-study-about-how-bad-ai-is-for-our-brains-183418494.html" rel="noopener noreferrer"&gt;There's yet another study about how bad AI is for our brains&lt;/a&gt; — Engadget&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://ktla.com/news/nationworld/ai-use-reduces-focus-persistence-study/amp/" rel="noopener noreferrer"&gt;Using AI for just 10 minutes can weaken cognitive skills, study warns&lt;/a&gt; — KTLA&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://knowledge.wharton.upenn.edu/article/without-guardrails-generative-ai-can-harm-education" rel="noopener noreferrer"&gt;Without Guardrails, Generative AI Can Harm Education&lt;/a&gt; — Knowledge at Wharton&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://the-decoder.com/openai-launches-study-mode-for-chatgpt-while-education-users-are-told-to-wait-and-learn-later/" rel="noopener noreferrer"&gt;OpenAI launches Study Mode for ChatGPT&lt;/a&gt; — The Decoder&lt;/li&gt;
&lt;li&gt;
&lt;a href="https://blog.google/outreach-initiatives/education/guided-learning/" rel="noopener noreferrer"&gt;Guided Learning in Gemini&lt;/a&gt; — Google&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;Read the full transcript, with every source, on &lt;a href="https://kuzscotech.github.io/kush-exe/ten-minutes-of-chatgpt-made-people-quit-sooner-study-finds/" rel="noopener noreferrer"&gt;the page for this video&lt;/a&gt;. More explainers like this one are on the &lt;a href="https://www.youtube.com/@kushdotexe" rel="noopener noreferrer"&gt;Kush_EXE YouTube channel&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;This article is built from the video's script and published sources. The video uses AI illustrations.&lt;/p&gt;

</description>
      <category>ai</category>
      <category>news</category>
      <category>programming</category>
    </item>
  </channel>
</rss>
