Privacy has always mattered and will matter even more in the AI age.
But there are people arguing that we live in a post-privacy world.
They say that people do not care about data privacy as long as the product delivers for them.
And on some level, they are probably right.
With so many digital products, data is collected by default, including in ways that are not immediately obvious. When you visit a website or use an app, a whole trove of data is generated and collected about you; what you do on the website or app, the other apps and websites you use, your location, and other bits of information constituting your digital footprint.
This data are used to target you with ads, sold to data brokers and even used to train the algorithms that decide what you see, how much you pay, and sometimes whether you qualify for a service at all.
This processing happens in the background. It consists of covert operations hidden away from the frontend experience that users are predominantly exposed to. What goes on under the hood has never really been their concern.
But AI changes this user experience significantly.
Think about how ads are deployed on a social media platform vs ads deployed inside an AI chatbot.
On social media, ads are mainly based on the data users provide passively. Platforms make inferences about their interests based on user activity and predict the products and services that are most aligned with those interests.
Inside an AI chatbot, the experience is far more intimate. Heavy users treat these systems as trusted companions: they bounce ideas off them, confide their private thoughts and feelings, and use them to organise their lives.
That intimacy makes advertising feel different. The ads are generated from the user’s own conversations and appear inside the thread itself, woven into an exchange the user experienced as private. They land as alien intrusions, disruptive in themselves, but also revealing. Each one is a reminder that someone was listening all along, and that the conversation was never really private.
Privacy stops being an abstract worry and becomes an immediate one, in a way few other digital products manage to provoke. The ‘off’ switch for personalised ads no longer feels like a preference buried in settings. It feels like a necessary act of reclamation, a way to restore some quiet to a space that was supposed to be yours.
And accordingly, consumers are unlikely to adopt AI products and services that do not provide the necessary privacy-preserving measures.
This will be the same for ambient AI agents that work in the background autonomously.
In this context, AI is being pedestalised as the intelligence engine sitting between all your different stores of data. If you are a consumer, it is your emails, messages, calendar, bank account.
Nobody hands over that kind of access without reassurance. The moment you start connecting data sources, whether onboarding an agent or building a workflow, three questions surface immediately: Will this data be handled reliably? Will confidential information stay confidential? And what recourse do I have when something goes wrong? Products that cannot answer these questions convincingly will not earn users’ trust.
But this requirement will not just be the expectation of individual consumers - enterprises will also have this on their checklist before procuring an AI system.
Before a business uses an AI product or service, they want to ensure that their data are not used for model training, that there is limited or zero data retention, that data residency controls exist and that the confidentiality of communications will be maintained, among other things.
Privacy will no longer be something that can be ignored, treated as a nice aesthetic, or bolted on in a rush. It will become a primary expectation, and AI providers who have not built it into their products will lose out. Consumers and enterprises simply will not use AI that handles data carelessly.
We are not in a post-privacy era. We are heading toward the opposite.