https://media.privacyinternational.org/videos/watch/7995fc21-039b-4305-9d32-3fae55c92d80 Every time you use ChatGPT, your conversations could end up shaping responses to complete strangers and giving OpenAI a window into your entire digital life. Here are 6 settings worth changing now to protect your data. AI chatbots are used by people in a variety of ways, including to conduct research, create content and analyses, deliver projects, and engage in conversations. We are concerned these AI tools can pose privacy and security risks when offered by tech firms, as they give these companies greater access to your data and access to our personal and professional lives. Prevent use of your data to train the model Large Language Models are trained on massive datasets. We have expressed concerns about the lawfulness of using data collected online for this purpose and there are demonstrable risks of this data being reproduced by AI chatbots without user consent or knowledge (called “regurgitation”). To limit the risk of personal data leaking out, we suggest preventing the content of your chats from being used for further training. On mobile: Go to your profile by clicking the menu in the top left corner and tapping on your profile in the bottom left corner. Navigate to Data Controls and turn off Improve the model for everyone. On browser: Go to the settings by clicking on your profile in the bottom left corner and clicking Settings. Navigate to Data Controls and turn off Improve the model for everyone. Review third party apps that ChatGPT can access ChatGPT can access documents, calendar events and other information from connected apps, such as cloud file storage. We have expressed concerns about the privacy and security risks created by connecting AI chatbots to other apps as it can be exploited to access personal data. To limit these risks, regularly review which apps ChatGPT is connected to and remove services you don’t use. Click your profile in the bottom left corner and click Settings. Navigate to Connected apps and review the apps you are connected to. Once this is done, go through the settings of the app you had connected and remove ChatGPT from the third party apps with access. Note that this feature is only available on your browser and not in app. Managing if and how your chats are shared ChatGPT allows you to share conversations via a URL. While in theory only people with the link can read the conversation, it’s not impossible for this link to end up online and be indexed by search engines. You can manage your previously shared chats in the settings by clicking on your profile and Data Controls. Click Manage next to Shared links and delete any chat you don’t want to be public anymore. Note that this feature is only available on your browser and not in app. Delete chats you won't come back to. Chats, like emails or messages, might contain sensitive information you don’t want others to access. To limit the risks in case of unauthorised access by someone with access to your device or a malicious actor, it can be good practice to delete chats you won’t come back to. Tap on a chat and click the three dots in the top right corner. Tap delete. You can also delete all your chats in the Data Controls menu. Enable Two-Factor authentication Two-Factor authentication provides an extra security step in order to access your account. This reduces the chances of someone accessing your account and getting access to your previous chats. You will need a dedicated app to setup Two-Factor Authentication. On mobile: Go to your profile by clicking the menu in the top left corner and tapping on your profile in the bottom left corner. Navigate to Security and tap multi-factor verification. A browser window will open and you will be asked to login. Follow the instructions. On browser: Go to the settings by clicking on your profile in the bottom left corner and clicking Settings. Navigate to Security and click multi-factor verification. You will be redirected to a new page, follow the instructions. Turn off ad personalisation ChatGPT has started to introduce ads in some regions, along with controls that let you limit how much your activity shapes what you see. For now this feature is only available to users in the US as part of a limited rollout. There is a setting to turn off ad personalisation and reduce how much of your data is fed into this targeting. This feature will be coming to other regions. When it appears, you’ll find it under Settings ; Ad Controls. If you turn ad personalisation off you'll still see ads, but based on the context of your current chat thread and not based on accumulated data from past conversations. #ChatGPT #privacy #AI #chat #chats #GPT #OpenAI #SamAltman #LLM #security #regurgitation #data #information #technology #tech #AIAgent #confidentiality #PersonalInformation #links #SearchEngine #twofactorauthentication #AIData #settings #OnlinePrivacy #LargeLanguageModel #ArtificialIntelligence Sun, 02 Aug 2026 20:23:43 GMT https://validator.w3.org/feed/docs/rss2.html PeerTube - https://media.privacyinternational.org https://media.privacyinternational.org/lazy-static/avatars/5c7454e0-676b-43a0-8933-4867d0c1f2e6.png https://media.privacyinternational.org/videos/watch/7995fc21-039b-4305-9d32-3fae55c92d80 All rights reserved, unless otherwise specified in the terms specified at https://media.privacyinternational.org/about and potential licenses granted by each content's rightholder.