AI Profiling Explained: Emotion, Movement & Mood Tracking

In the first post in this series, "Your Data Is Listening and Watching", we explained how every click, swipe and voice command gets logged and stitched into an invisible profile that algorithms use to influence what you see.

In the second, "Cambridge Analytica Rebranded", we showed how those psychological targeting tools didn't vanish. They just reappeared under new names and a fresh coat of paint.

Now, in this, our third post in the series, we’re fast-forwarding to 2025—where profiling isn’t reactive or static. It’s real-time, relentless, and happening whether you notice it or not.

Forget surveys and "take this quiz". Today's AI watches how you speak, move and even breathe, then builds a profile before you can blink.

This post explains how your phone, smartwatch, car and smart home sensors are all "phoning home", feeding AI systems that build emotional and behavioural snapshots of you. That includes the smart home packages, security cameras and app-controlled devices many builders now offer in a new home.

We'll go from old-school psychometrics to always-on surveillance, show how your data becomes someone's ad strategy, and explain why that should worry you, or at least make you want your rights back.

How AI Builds You: From Manual Profiling to Real-Time AI Insights

Picture the difference between sitting in a quiet room filling out a personality test, and going about your day while AI quietly pieces together your emotional state. That's the gap between old-school Cambridge Analytica-style profiling and today's real-time AI profiling.

Manual Profiling: Solid, but Old-School

Profiling used to mean questionnaires or standardised tests. You consciously reported how you felt or what you preferred. It was explicit, but slow and stuck in fixed categories: think MBTI (Myers-Briggs Type Indicator), demographics or psychometric buckets.

Real-Time AI Profiling: Always On, Always Learning

Fast forward to today. AI doesn't wait for your answers; it observes you. Through phone apps, wearables, connected cars, smart home gadgets and the software you use, it collects a constant stream of behavioural signals:

Wearables and body signals: Smartwatches don't just count steps. They can infer stress or anxiety in real time from skin conductance or heart rate variability.
Emotion sensing: Research systems can already combine camera, microphone and sensor data to guess your mood, in places like smart homes and healthcare.
App and device telemetry: Ever noticed your phone or car seems to "phone home"? Telemetry records your habits, from when you open an app to how long you look at a screen, and feeds AI models that spot patterns at scale.
Connected devices everywhere: Your fridge, home sensors and car audio system all generate data. AI can analyse these signals and infer mood or behaviour without you realising.

With that much data streaming in, AI can:

Detect emotion in real time (e.g. joy, stress, anger) without you telling it
Group people dynamically by behaviour, time of use, mood and context, not fixed personality types
Scale instantly, profiling huge numbers of people as the data pours in

From Data Collection to Commercial Targeting

Once AI has grouped you, businesses and plenty of others line up to use that information. The ads you see aren't just based on what you searched or bought, but on when you're distracted, relaxed, anxious or primed to spend.

How it works (generally speaking):

Your phone, smartwatch or car picks up small behaviours.
AI groups you with people who behave similarly: "late-night browsers", "stressed commuters", "keen weekend exercisers".
Those groups are sold or used across marketing systems to tailor offers and ads, or to influence you.

It all happens in the background, with no forms and no meaningful consent. You don't earn a cent. You probably paid for the device that collects it.

Ethical & Privacy Risks

As AI gets smarter and more independent, the privacy worries stack up. Take agentic AI, which acts with little human oversight. Efficient, sure, but when it gets it wrong, who's accountable?

One case to watch is below. It involves the death of a teenager. If this raises anything for you, call Lifeline on 13 11 14.

This case is before the courts. How it's decided will show how far AI companies can be held accountable.

The Stanford 2025 AI Index Report (below) found a 56% jump in reported AI-related incidents in a single year.

The very systems profiling us in real time aren't just watching; they're also exposed to mistakes, misuse and bad actors.

Regulation & Governance Landscape

Guardrails are finally being bolted on. The EU AI Act came into force in August 2024, with its bans (including social scoring and most real-time facial recognition by police in public places) applying from February 2025. It also demands extra transparency for high-risk AI systems.

The UK followed with its Data (Use and Access) Act 2025, which allows some automated decisions, but only with human intervention available, a right to challenge, and a process that isn't a total black box.

Closer to home, Australia passed the Privacy and Other Legislation Amendment Act 2024 in late 2024, including a new statutory tort for serious invasions of privacy, with a second tranche of privacy reforms still pending. The OAIC has the details.

💡

But let’s be real: laws move at a snail’s pace, while tech sprints. So don’t hold your breath waiting for regulation to catch up any time soon.

Conclusion

AI companies are in an all-out sprint, hoovering up data to beat the competition. In that mad dash something gets trashed: your rights, your privacy and, bit by bit, the social fabric we all rely on.

With that power comes a responsibility that's too often missing. Where's the consent?

*The oversight? The fairness?* This isn't just about slicker marketing. It's about protecting your autonomy when the commercial gaze never blinks.

Look at the toll social media has taken on young people. Internal TikTok documents (read more here) allege the app can become habit-forming in as little as 35 minutes, and that heavy use is linked to anxiety, memory problems and reduced empathy.

Unsealed internal videos show TikTok staff raising concerns about the algorithm pushing disordered-eating content to teens. In France, seven families have sued, claiming TikTok's algorithm exposed their children to self-harm content.

These stories aren’t isolated. They're a warning: the race for growth and to embrace the new often tramples real human costs.

So: will we learn from this, or keep sprinting towards the future and spend the next era cleaning up after unchecked AI profiling?

💡

If this sounds concerning, you're not alone. Speak up—demand clear privacy policies from your apps and devices. Support legislation that protects against opaque surveillance. And, most importantly, remain conscious of what you share—because in a world where your profile is being built behind the scenes, knowledge is your most powerful tool.