Privacy in the Age of AI: How Your Secrets Are Inferred, and How to Protect Them

Imagine a stranger guessing your mood from the tone of your voice, and your habits from the hour you go online. This is no longer fiction; it is the quiet reality we live in. The question is not “when will we lose our privacy?” but “have we already lost it, while signing terms we never read?”

From walls to patterns

Privacy used to mean walls, curtains and locks. A secret was a physical thing that could be hidden. Today the secret has become a pattern, the pattern has become data, and the data has become a commodity. In her book on “surveillance capitalism”, Shoshana Zuboff, professor emerita at Harvard Business School, describes how large companies extract a “behavioural surplus” from our conduct, build predictions of our actions on it, and seek to steer them. What is new is not that we are watched; surveillance is old. It is that surveillance now predicts and intervenes, and studies have shown that what can be inferred about us exceeds what we disclose.

Inference is more dangerous than disclosure

The danger no longer lies in what you reveal, but in what the system deduces without your saying a word. In 2013, Michal Kosinski and colleagues showed that Facebook “likes” alone were enough to predict personal traits of their owners, who had never declared them. What you deliberately conceal may be exposed by what you do not notice.

The “consent” no one reads

Traditional law was built on “informed consent”. But how can anyone consent to a mechanism they do not understand, or know that data given today will be combined years from now with other data to produce a picture that was impossible when they agreed? Consent has become a ritual: accept, or be shut out of digital life.

Nor is “anonymisation”, which some companies rely on, a sound defence. A study in Nature Communications (2019) estimated that fifteen demographic attributes suffice to re-identify 99.98% of Americans, even in incomplete, “anonymised” datasets.

Is the secret dead?

Not yet. The European General Data Protection Regulation (GDPR) was a turning point: it gave data subjects enforceable rights (access, rectification, erasure, portability and objection) and set fines of up to 20 million euros or 4% of global turnover. In the Kingdom, the Personal Data Protection Law was issued to balance innovation with safeguarding individuals’ rights. Yet regulation alone is not enough; law runs breathless behind technology, and is sometimes written in loose language.

Even the feeling of being watched changes behaviour and pushes people towards self-censorship (Penney, 2016), narrowing the room for trial and error in which freedom and creativity grow. The deepest threat to privacy is not the exposure of secrets but the drying up of the springs of a free self.

Four pillars of protection

1. Design and legislation: “privacy by design” as the default, not an afterthought.
2. Technology: advanced encryption, federated learning (McMahan et al., 2017) and differential privacy, which analyse data without exposing individuals.
3. Knowledge: awareness that data is an asset, not a by-product, and teaching the young that privacy is a right worth defending.
4. Institutions: effective regulators with real powers.

The “I have nothing to hide” fallacy

Some think privacy is the need of wrongdoers alone. In truth it is a condition of human dignity. A sound person closes the door of their room not because they are committing a sin, but because they have a sanctity and boundaries. Our heritage values concealment of faults (sitr) and the inviolability of homes, natural allies of a genuine philosophy of privacy; and the Quran gives an explicit prohibition of spying: “and do not spy” [Al-Hujurat: 12]. The fallacy shifts the burden of proof onto the victim, instead of requiring the watcher to justify the wish to look.

The secret has not died, but it has entered intensive care. Its fate depends on what we do today: will we be open books, or renegotiate the terms of our digital existence? Trustworthiness with data, and gentleness towards its owners, are what remain after technologies change.