Gen launches AI Data Checker to show how AI apps use your data

25 Sep 2026 (ET): Gen Digital opened a beta of AI Data Checker — enter an email or username and it lists connected AI apps plus what they collect, whether data may train models, retention, and privacy controls — as part of its Agent Trust Hub.

SOFTWARE desk — agents are starting to act with people’s accounts, from a purchase to a trip to a bill, and the rules for that data still live in long privacy policies. A readable map matters because a person can enter an email or username, see which AI apps are connected, and read what each one may collect, whether that information may train a model, how long it may be kept, and which privacy controls exist, before they hand an agent more of their life.

What the release says the product is. Gen (NASDAQ: GEN) launched a beta of AI Data Checker, built by the Gen AI Foundry. The page calls the Foundry the company’s hub for emerging AI innovation. A beta is a test version people can try before a finished release. That gloss is this desk’s. The checker is the latest tool in the Agent Trust Hub. Its job, on this page, is to make it easier to see how popular AI apps and agents handle a person’s data. The subhead says Gen data shows nearly one in three people use AI apps and agents, as the company expands the hub with this checker. The body uses a narrower count. Do not collapse those two sentences.

The telemetry the body cites. Telemetry, here, means Gen’s own measurements of what its desktop users do. That gloss is this desk’s. The release says nearly one in three desktop users now interact with agentic AI services. Agentic, here, means the software is being asked to take actions, not only to answer a question. That gloss is this desk’s. The opening says people are asking AI to research purchases, plan trips, manage everyday tasks, and handle finances. Nearly one in three is the company’s figure. This desk did not audit the measurement. The subhead’s “nearly 1/3 of people” and “AI apps and agents” are not the same words as “desktop users” and “agentic AI services.” Keep them apart.

How a person is supposed to use it. The release says you enter an email or a username. The checker then lists connected AI apps and services that person uses. It also scans non-AI accounts where the person may have connected and potentially shared personal information. For each service it shows what information or data the service collects, whether information may be used to improve or train AI models, how long information may be retained, and what privacy controls are available. Train, here, means the service may use that information to improve its model. Retention is how long the service says it may keep the information. Those glosses are this desk’s. The page says this is a clearer starting point, so a person does not have to read a long privacy policy or become an expert first. This desk did not submit an email or a username.

What the hub already had, in the release’s words. Gen says it created the Agent Trust Hub, shortened to ATH on the page, so people would have more visibility, understanding, and control when they use AI assistants and agents. Since launching in February, the hub has added VPN for Agents and Sage. A VPN, a virtual private network, is a tunnel that hides a device’s traffic from the network it is on. That gloss is this desk’s. The release calls VPN for Agents the industry’s first AI-native VPN purpose-built for autonomous agents. It calls Sage the first consumer agentic security layer, available through Norton AI Agent Protection in Norton 360. First, industry’s first, and February are the company’s words. The page does not print the year of that February. Do not add one. AI Data Checker, the release says, moves the hub from protecting how agents operate to showing the data relationship behind the tools a person already uses.

Howie Xu, as a quote, not as a lookup this desk ran. Xu, Chief AI and Innovation Officer at Gen, said: “AI is becoming much more personal because we're asking it to do more for us.” He said that can be incredibly useful, but people should understand what they are giving an AI access to in return. He said the AI Data Checker is about making that easier to see, so people can make choices about the AI tools they use with more confidence. That is his sentence on the wire.

What else the release names, and where the checker lives. The page says visibility has to sit next to protection. It names Norton Genie Scam Detector, Norton AI Agent Protection, and Deepfake Detection in the Norton and Avast brands, plus other products it does not list. A deepfake, here, is a fake image, video, or voice made to look real. That gloss is this desk’s. The new Norton Family Assistant, the release says, is a private AI assistant, supported by Norton technology, that can handle tasks with your permission and manage your digital life. Gen also invites people to sign up for early access to later tools. The checker is available now at https://ai.gendigital.com/, and the same address is where the release says you can sign up. “Live fearlessly in the AI world” is Gen’s line on that sentence. It is not this desk’s. The media contact on the wire is Brittany Posey, Gen, Press@GenDigital.com.

Who Gen says it is. The about box calls Gen a global company for Digital Freedom, through Norton, Avast, LifeLock, MoneyLion, and more. It says those brands cover cybersecurity, online privacy, identity protection, and financial wellness, for millions of people in more than 150 countries. Millions, and more than 150 countries, are the wire’s. The Gen AI Foundry, on the same release, is described as the engine behind recent AI work: security experiences, autonomous financial helpers, and identity agents, plus the trust infrastructure that includes the Agent Trust Hub. The about box does not print the product page’s “500M+” figure. Do not move that number onto this release.

What the product page showed, kept apart from the wire. This desk loaded https://ai.gendigital.com/ on 25 Sep 2026. The title is “Agent Trust Hub | How is AI using your data?” The headline is “How is AI using your data?” The line under it says your chats can train AI and fuel targeted ads, and tells you to find out which apps use your data. The search field’s placeholder is “Email or username.” The button is “See my data.” A screen-reader label on that field says “Your email address.” The note under the button says that by continuing you confirm you are 18 or older, authorize the search, and acknowledge the privacy notice. A trust line says 500M+ people protected, 40+ years of protecting consumers, and “From Gen, the Company Behind,” next to a Norton mark. 500 million-plus and 40-plus years are this page’s. They are not in the wire’s about box. The page also has a block that says “Your privacy has an agent now,” with “Get early access!” and a field whose placeholder is “Add your email address here.” That early-access field is this page’s. The wire points early access at the same site and does not print that placeholder. The nav names Norton Family Assistant, Sage, Skill Scanner, and Agent VPN. Norton Family Assistant on that nav points at us.norton.com/products/family-assistant. The wire names VPN for Agents and Sage. It does not name Skill Scanner. The site’s link text is “Agent VPN,” at /agentvpn, not the wire’s “VPN for Agents.” Do not collapse those names. This desk did not run a search.

How the page says it grades an app. The wire does not print a scoring system. The page’s FAQ says it reads the app’s privacy policy and terms and answers 15 fixed questions, the same questions for every app. For most apps those questions are the whole score. If an app can buy, book, click, or send for you, 90 points stay on the policy and 10 points ask whether its own page says it refuses an unsafe website. If it does not, those 10 points are zero, and the report says it may transact on unsafe websites on your behalf. Every policy answer, the page says, must quote a line it checks word for word. The letters: A+ from 64, A from 56 to 63, B from 48 to 55, C from 40 to 47, D from 34 to 39, and F below 34. On the tier list, A+ is shown as S. No privacy policy at all means an automatic F. A full report, the page says, expands “How we got here” into all 15 answers, each with a red, amber, or green dot and the quote behind it. Those bands are the page’s method. This desk did not score an app.

The Instinct card on that page, kept apart from the wire. The hero preview is Instinct, labeled Highest risk, under the line “From its privacy policy.” The quoted policy line is: “Even if you opt out, we may still use your information for AI model training when that information is flagged for safety review.” The sentence under the quote says turning off training does not stop them from using some of your chats to train safety systems. A link says “Read the full report.” A disclaimer says Agent Trust Hub uses AI and may make mistakes, and that you should review a report and confirm it before you rely on it. A carousel under the hero labels other named apps with lines such as Highest risk, Medium risk, Low risk, or a U.S. security advisory. Those labels are the page’s. This filing does not treat them as findings this desk checked against each policy.

What the card shows. The card is a headless capture of that product page: the Agent Trust Hub search, the “See my data” button, and the Instinct Highest-risk preview, on a navy margin. It is not the wire’s graphic. That wire image is captioned “Gen releases AI Data Checker to help people understand how AI handles their data,” and the file name on the wire begins ChatGPT-Image and includes a clock. This desk did not use it. The card has no desk date painted on it.

PRIMARY here: Gen Digital Inc.’s 25 Sep 2026 PR Newswire release 302890485, visible stamp Sep 25, 2026, 14:02 ET, datelined Mountain View, Calif. — Tier A PRIMARY, the company’s own wire. STATUS PRIMARY. The desk label on the chip is CONFIRMED because that is the catalog word for a primary page we can stand on. The beta, the Gen AI Foundry, the Agent Trust Hub, the nearly one in three desktop users, the email or username lookup, the connected-app list, the non-AI account scan, the four data fields, VPN for Agents, Sage, the February launch with no year printed, the Xu quote, the named protection products, Norton Family Assistant, the early-access invitation, and the about box are that wire’s. The product page’s placeholder, the 500M+ and 40+ lines, the 15-question grade bands, the Instinct preview, and the carousel labels are that page’s, read with this filing. They are not a second announcement. The catalog chip is SOFTWARE, not SIGNAL.