[{"data":1,"prerenderedAt":28},["ShallowReactive",2],{"nr-en-grok-built-apps-expose-patient-data-and-plaintext-passwords":3},{"slug":4,"title":5,"dek":6,"date":7,"time":8,"publishedAt":9,"updated":10,"updatedAt":10,"dateFmt":11,"updatedFmt":10,"kind":12,"tier":13,"author":14,"authorName":15,"topics":16,"tracker":10,"trackerLabel":10,"headlineStat":22,"image":23,"ogImage":24,"imageAlt":5,"csv":10,"minutes":25,"words":26,"html":27},"grok-built-apps-expose-patient-data-and-plaintext-passwords","Grok-built apps expose patient data and plaintext passwords","A security firm found that nearly 95 percent of more than 37,000 web apps built with Grok were accessible without login. The exposed data included patient lab values, confidential documents from SpaceXSI and X, and plaintext passwords.","2026-11-11","14:55","2026-11-11T14:55:00+01:00","","November 11, 2026","news","standard","ideal-syka","Ideal Syka",[17,18,19,20,21],"AI security","Data protection","Grok","Patient data","AI-generated code","94.7% of 37,571 apps openly accessible","\u002Fnewsroom\u002Fimg\u002Fgrok-built-apps-expose-patient-data-and-plaintext-passwords.webp","\u002Fog-nr\u002Fgrok-built-apps-expose-patient-data-and-plaintext-passwords.en.png",3,615,"\u003Ch2>What happened\u003C\u002Fh2>\n\u003Cp>According to security firm Claroc&#39;s report, it checked 37,571 apps that were still online. Of these, 35,578 were open to anyone, or 94.7 percent. You can infer how widespread the problem may be for AI-generated apps, although the report concerns Grok apps specifically.\u003C\u002Fp>\n\u003Cp>Claroc traces the cause to the publish dialog. On the paid test account the firm used, &quot;Anyone with the link&quot; was already preselected. Each published app gets its own subdomain, which lands in searchable passive DNS databases after the first visit and can be found by pattern.\u003C\u002Fp>\n\u003Ch2>Exposed data\u003C\u002Fh2>\n\u003Cp>According to Claroc, 24 of the accessible apps came from SpaceXSI and X employees. They included a confidential onboarding plan with internal system addresses and 19 chat channels, a publicly accessible sales territory map with 594 strategic accounts, and eight X ad proposals, one of which targeted a U.S. federal agency and was marked confidential.\u003C\u002Fp>\n\u003Cp>Most of the open apps were games, landing pages, or calculators, Claroc says. But about one in ten contained sensitive material, internal business data, or personal information. The researchers estimate 650 to 800 apps contained private data. Confirmed cases included an ICU briefing with a patient&#39;s lab values, a list of more than 3,500 employees at a hospital group, and a fire department shift schedule. The source code of a timesheet app contained 18 accounts with plaintext passwords.\u003C\u002Fp>\n\u003Cp>Claroc also identified 30 affected organizations, six of them Fortune 500 companies. The firm is withholding their names.\u003C\u002Fp>\n\u003Ch2>Flawed access controls in test apps\u003C\u002Fh2>\n\u003Cp>Two of Claroc&#39;s own test apps showed another problem. A contact list app had a database with no login, so anonymous visitors could add entries. For a sales team customer list, Grok itself added Google and X sign-in. However, the Grok-written code added every authenticated user to the team first and only checked membership afterward. An outside Google account could therefore read, edit, and delete records. Claroc says it ran write and delete tests only on its own apps.\u003C\u002Fp>\n\u003Ch2>Response and demands\u003C\u002Fh2>\n\u003Cp>SpaceXSI responded, according to the report, that apps are private until the creator shares them, that publishing involves several steps including a legal notice, and that employee accounts default to &quot;Only you&quot;. Claroc counters that extra clicks don&#39;t help if &quot;Anyone with the link&quot; is still preselected at the final step. The 24 open employee apps also show how little the stricter staff default achieves.\u003C\u002Fp>\n\u003Cp>Claroc proposes making &quot;Only you&quot; the default and showing who will have access before publishing. Grok should also warn when an app has a database but no login. Instead of individual subdomains, apps should run on hard-to-guess paths on a shared host. According to the report, SpaceXSI has taken the 24 employee apps offline but has not removed the other open apps from users and organizations.\u003C\u002Fp>\n\u003Cp>Claroc has offered to share the full list of addresses confidentially. Because SpaceXSI did not commit to contacting affected owners, Claroc is notifying them directly. The firm sells security solutions for AI-built apps and therefore has its own interest in the topic.\u003C\u002Fp>\n\u003Ch2>Assessment\u003C\u002Fh2>\n\u003Cp>The report is at least the third documented privacy incident involving SpaceXSI and Grok, after more than 370,000 shared Grok conversations appeared in Google Search in August 2025 and the Grok Build terminal tool uploaded user files in July 2026. For German companies using AI tools to build internal apps, default publishing settings may matter a great deal. Whether this transfers to their own projects should be checked case by case; the figures come from a vendor report written by Claroc itself.\u003C\u002Fp>\n\u003Ch2>Sources\u003C\u002Fh2>\n\u003Cul>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fthe-decoder.com\u002Fgrok-built-apps-exposed-patient-data-plaintext-passwords-and-spacexsis-own-sales-pipeline\u002F\">The Decoder\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>\u003Cem>Editorially owned by \u003Ca href=\"\u002Fen\u002Fautor\u002Fideal-syka\">Ideal Syka\u003C\u002Fa>. Sources and method: \u003Ca href=\"\u002Fen\u002Fredaktion\">Newsroom &amp; method\u003C\u002Fa>. Tips and corrections: \u003Ca href=\"mailto:ai@i6eal.de\">ai@i6eal.de\u003C\u002Fa>.\u003C\u002Fem>\u003C\u002Fp>\n",1794482128130]