AI Privacy Guide helps beginners protect personal data, understand AI risks, and stay secure in the digital world with easy privacy practices.
Artificial intelligence assistants ask users to activate some features that should be optional. These are the mandatory ones and the ones that should not be enabled.
Table of Contents
These data contrast with those shared just a few days ago by OpenAI, developer of the most widely used chatbot globally; This is endorsed by its almost 700 million weekly users, a figure that gives an idea of how this technology has been introduced into people’s daily lives.
AI is unstoppable, as it not only solves some mechanical and tedious tasks but can also offer new ideas about how to approach the analysis of a text or is even able to summarize what has happened in a meeting, as well as give advice on how to handle certain situations.
Although in some cases it is not recommended to use it (for example, using it as a therapist to solve a personal or mental health-related problem), in others it can be a very useful and interesting tool. So much so that it is now much easier to plan a trip with her than to do it step by step, the old-fashioned way.
In all cases, however, we will be giving this technology very personal information, and sometimes we do not know what its final destination will be and if it is really necessary for the AI to know such intimate and sensitive data. So, the question is, how far can artificial intelligence go?
Actually, as far as we can go, as long as we carefully review the permissions we grant to these chatbots and assistants before launching them. Based on this, it is worth remembering which permissions an AI should request as a requirement to use its service and which ones it should offer as optional.
User account and internet access
As with any online tool, the most important thing of all is the simplest: carefully read what is being offered to us and what we are accepting or denying. And, in this sense, the first thing we must choose is whether to use the free version of a certain assistant or opt for one of the available paid plans.
Once we have done so – yes, believe it or not, it is very common to get blocked at this point, because brands try to do it until the end to get you to make a paid account on their platforms—it is most likely that there will be a deployment of tabs with functionalities to which you have to give express consent to be able to use them.
Among the permissions that are considered mandatory is that of access to the internet, which is essential for artificial intelligence to function, because this technology must communicate with external servers to process information, as well as to access the language model chosen for its operation.
It is also common to request access to the user account if you have previously opted for a paid version or if you already have an account created in its service. This is the case with Gemini with Google or Copilot, which offers a personalized experience to users with a Microsoft account, with the Work and Personal alternatives.
This is completely optional and can be used both with and without an account in the free mode, although it is generally common to log in if you already have a user account due to the benefits this offers. For example, the storage of the history of requests or the personalization of responses based on previously made queries.
Watch out for GPS location or browsing history
The vast majority of attendees justify access to personal data to offer a better experience despite the fact that, in reality, they do not need it to perform their work, since they should be able to respond to the vast majority of requests without using certain information.
And there are some platforms that do it deliberately, without clearly warning those who use them, stressing that they do so because they already explicitly point it out in the Privacy Policy section that, to tell the truth, most users do not read. This is the case of ChatGPT.
Other artificial intelligence services, such as Gemini, anticipate that it is possible to enable the option with which personal information is shared for the training of third-party models, although it also insists that if this platform is used to interact with these others, the data will be treated “in accordance with their own privacy policies.”
In any case, when using a chatbot, it is important to be very careful about enabling permissions that, a priori, are not necessary for AI. For example, access to the camera, which is only essential if you are going to use visual recognition functions or video calls. Also the microphone, which is not necessary if you are not going to make requests by voice.
It should not be mandatory, although optional, to share location or enable geolocation services, unless artificial intelligence needs it to fulfill its function. For example, to recommend nearby restaurants or to find a nearby gas station.
It is also a serious violation of privacy for AI to have access to browsing history, as this technology does not need to know the searches a user has made to give them a certain answer. Therefore, it should not be mandatory to grant him that permission, nor is it legitimate for him to ask to know the contact list.
Based on all this, the important thing is not only to know what personal data of ours an artificial intelligence can know, but also at what time it should do so. Therefore, it is advisable to disable functions that are not needed in daily use to minimize risks and frequently review your privacy settings.
Be a part of over Success!
- Stay ahead of the curve with the latest tech trends, gadgets, and innovations! Newsletter
- Follow me on Medium for more insights
- Write for Us on Technoluting (Medium)
- Share your feedback or connect with me on LinkedIn — I’d love to hear from you!
