Google’s new AI could be a privacy nightmare

Apparently, take over the entire open Internet was not enough.

In an attempt to both score points in its competition with iMessage and in its race in the development of AI, Google announced the deployment of the Bard conversational assistant in its Messages application on all Android devices . If this announcement is not really surprising, and even logical, the American company has once again fallen into its bad habits, and took advantage of the opportunity to create a new method of collecting your information. And this time, it’s your private discussions that are directly targeted. Let’s see how long the EU will take before issuing fines.

Google includes Bard in its Messages app…

Bard himself says it: its introduction into the Messages app will be a “game changer”. The conversational assistant promises to “facilitate communication, develop creativity, and provide information”but also “help you write messages, translate languages, identify images…”. In addition to providing you with additional options when chatting with your contacts, you will be able to have access to Bard directly from the application. The addition does not seem absurd.

And from Google’s point of view, it allows both to compete with iMessage, since Apple seems a little behind in the development of artificial intelligence, and to catch up on ChatGPT by taking advantage of its enormous strength in distribution, which OpenAI probably can’t compete with.

…and your messages in Bard

But of course there is a wolf. The messages you write directly to Bard can be – and likely will be – used to train the model, but it’s ultimately the same logic as the queries you can make to ChatGPT. Since this is equivalent to sending a message to Google directly, this part is pretty understandable.

But according to Bard, in order to provide you with the best service and the most personalized experience possible, the chatbot will also analyze the content of your private messages. “to understand the context of your conversations, your tone, and your interests”. Furthermore, your little helper has also planned “analyze your conversation history with your different contacts to understand the dynamics that take place in your different relationships, in order to personalize responses depending on the person you are speaking to”.

Google, which does not really have the reputation of being exemplary on the issue, is therefore breaking down all privacy protection barriers. And while Bard doesn’t explicitly say that your data will be used to train its artificial intelligence models… it doesn’t formally rule it out either. The giant, after having swallowed up the entire known Internet to try to regain the lead in the race for artificial intelligence, seems to not want to stop.

Such a level of violation of privacy, if it is really implemented as it stands, should give ever more work to the company’s lawyers, because there is little chance that the various control bodies, national governments or that of the EU which has recently implemented concrete actions, are letting this pass without taking a closer look.

Source : Forbes