Cameras are everywhere – from smart doorbells on front porches to quiet domes on grocery store ceilings – but do we really know what they’re doing? The outward appearance of a camera gives no hint of its true super powers. And the expectations of what cameras are able to do, lag behind their actual capabilities. Reasonable expectations, that are built over time, are no longer reflecting reality. This gap between what technology does and what we expect it to do is something I really first started to explore when writing the Book “Foundations and Opportunities of Biometrics”. 

A surveillance camera in a grocery aisle might be completely inactive, or it could be passively recording footage, actively monitored by security staff, analyzing shoppers’ behaviors in real time, or even running facial recognition on every face that passes by – still looking like the types of cameras we’ve seen being there for decades.. 

The same goes for a doorbell camera: it may simply record video, or it might also capture audio, detect motion and people, identify familiar faces, and upload everything to the cloud. Yet, when you glance at these devices, nothing about their appearance reveals their super powers. This gap between what cameras look like and what cameras actually do is a growing problem for privacy, ethics, and informed consent. 

The Problem: Cameras with Secret Capabilities

Modern cameras have become mini-computers with lenses, with access to almost unlimited computing power from the cloud. Thanks to cheap sensors, internet connectivity, and AI, a basic-looking camera can pack an array of features. Consider a grocery store security camera. It could be:

• Off or Dummy: installed for deterrence but not actually recording.

• Recording Only: capturing video to storage, to be reviewed if needed.

• Live Monitored: streaming to a security room where someone watches in real time.

• AI-Analyzed: using software to count people or flag suspicious behavior.

• Facial Recognition Enabled: identifying customers (e.g., known shoplifters or VIP shoppers) as they walk in.

Crucially, you cannot tell which of these is happening just by looking at the camera. There have been real cases underscoring this invisibility. For example, the Rite Aid pharmacy chain quietly added facial-recognition systems to 200 stores across the U.S. Shoppers had no easy way to know that a simple ceiling camera was scanning their faces. Rite Aid claimed it posted signs about this surveillance, but reporters found no notice in over a third of the stores using the technology. In other words, many customers were unknowingly subjected to AI-powered identification as they did their shopping.

Now think about consumer doorbell cameras. A device like Google’s Nest Hello looks like a normal doorbell with a tiny camera, yet it can perform facial recognition – it literally recognizes who’s at your door and can tell the homeowner via smartphone . Google markets this as a convenient feature (“know if it’s the kids or a stranger at the door”), but from a visitor’s perspective, you’d have no clue that the doorbell you just pressed is analyzing your face against a database. Other doorbell cams, like Amazon’s Ring, can record audio, use night vision, and share footage to cloud servers – again, capabilities invisible to the casual observer. The public space is increasingly peppered with such smart cameras whose capabilities are hidden. 

This lack of transparency creates an information asymmetry: camera owners and manufacturers know exactly what their devices are doing, but the people being observed do not. It’s a bit like serving someone food with no ingredient list. You might be feeding on more than you bargained for. This is where the idea of “camera-ingredients” comes in – a call for greater transparency about what any given camera is capable of.

Imagine how many times a day a food delivery driver or mail/parcel delivery person is captured on a camera and potentially recorded and identified. 

“Camera-Ingredients”: A Transparency Metaphor

The term camera-ingredients is inspired by nutrition labels on food items like breakfast cereals. Just as a nutrition label lists what’s inside your food, a camera-ingredients label would clearly list what a camera can do. 

Why do this? Because in privacy and data protection law (like Europe’s GDPR), transparency is paramount – people have the right to know when their personal data is being collected and how. In practice, though, current camera notices are often just small signs saying “CCTV in operation.” That’s about as informative as a food label that simply says “Edible contents inside” with no details.

Under GDPR, organizations must inform people that they’re under surveillance, and guidelines even say the notice should include any information that might “surprise” the data subject – for example, if footage is shared with third parties or kept for a long time. In the UK, a standard CCTV sign will typically name who’s operating the system and a contact number, and maybe the basic purpose (“for safety and crime prevention”). But it won’t tell you the technology specifics. Are the images analyzed by AI? Is there audio recording? How long is footage kept? These details are usually buried in privacy policies (the second “layer” of information) or not disclosed at all. In effect, the public is kept in the dark about the “ingredients” of surveillance – the data and processing behind the camera’s eye.

The camera-ingredients concept proposes a simple, standardized disclosure that travels with the camera, much like a food label on a package. Imagine a sticker or sign right below a doorbell camera or on the wall of a store entrance, with a concise list of that camera’s capabilities. 

It might say things like “Live Monitoring: Yes”, “Facial Recognition: No”, etc., in plain language. This would transform the current vague warning signs into something more informative and empowering for the public. Just like the nutrition label on cereal boxes, it’s not perfect, but it makes it possible for people to understand key elements of what they contain.

In fact, a survey found that 92% of people believe it’s important to inform consumers when personal data is being collected , yet today it’s nearly impossible to get that information at a glance. A camera-ingredients label directly addresses this gap by making transparency tangible.

What Would a “Camera-Ingredients” Label Look Like?

How exactly do we disclose a camera’s capabilities in a clear, standardized way? The good news is we have models to draw from. Nutrition labels on food are a familiar template – they condense complex information (vitamins, percentages, ingredients) into a format anyone can quickly scan. 

We’re also seeing early versions of privacy labels for tech: researchers at Carnegie Mellon University recently developed a prototype “security & privacy nutrition label” for IoT devices . It’s a simple summary of what data a device collects, how it’s used, and with whom it’s shared, much like a privacy factsheet. The goal is to make privacy features easy to compare, just as one might compare calories on cereal boxes. In fact, their label for a smart camera includes details on data practices in a user-friendly format. This example shows that it’s feasible to convey things like what sensors a camera has (video, audio, etc.), where data is stored, and who can access it – all in a compact label design.

Such a camera-ingredients label could be affixed on the device or posted publicly where the camera operates. For instance, a doorbell camera might come with a sticker that the homeowner is encouraged to display for visitors. A business might put a standardized sign at the entrance listing the capabilities of their surveillance system. The label would focus on the capabilities that matter most for privacy and ethics. Here’s an illustrative example of what a camera-ingredients label might list:

Camera Feature	Enabled?
Live Monitoring (24/7)	Yes
Video Recording	Yes (stored 30 days)
Audio Recording	Yes
Night Vision (Infrared)	Yes
Motion/Person Detection	Yes
Facial Recognition	No
Shares Footage Externally	No

Table: An example “camera-ingredients” label for a hypothetical camera system, clearly stating its capabilities.

In this mock label, anyone can instantly see what the camera can and cannot do. For example, it’s doing live monitoring and recording video with a 30-day retention, it records audio, and it has AI-based motion detection – but it does not perform facial recognition and does not share footage outside the organization. Compare this to the vague “CCTV in use” sign we see today; the difference is night and day in terms of clarity. The label could even include a QR code or web link to a more detailed privacy policy (a “second layer” of information), similar to how GDPR envisions a two-layer approach to privacy notices . But the key is that the primary layer – the camera-ingredients – is front and center, concise and standardized.

By standardizing the format and wording (much like nutritional facts are standardized), people wouldn’t need special technical knowledge to interpret it. You’d quickly learn the icons or terms (for instance, a standard icon for facial recognition or a crossed-out icon if it’s off). This could be enforced or encouraged through regulation: just as food companies are required to display nutrition facts, perhaps one day camera deployments in public spaces could be required to display surveillance facts. 

Conclusion: Toward a More Transparent Surveillance Culture

We are living in an age where cameras are not “just” passively watching, but actively computing – identifying, analyzing, and cross-referencing data about us. In such a world, maintaining the status quo of minimal disclosure is a recipe for abuse and public backlash. Embracing the concept of camera-ingredients labels could help rebuild trust and accountability. It aligns with the spirit of laws like GDPR that seek to put individuals in control of their personal data by ensuring they’re informed. Transparency won’t solve all concerns about surveillance (we still need limits on what cameras should do, even if disclosed), but it’s a crucial first step.

Imagine walking into a store and seeing a sign that reads something like: “This premises uses smart CCTV (Live monitored, records audio; no facial recognition). Data kept 14 days for security purposes.” 

Such a notice treats people with respect – it tells you exactly what “ingredients” of monitoring you’ll be subject to, allowing you to make an informed choice about entering or perhaps to voice your concerns. It also holds the camera operators accountable to their word (if they say “Facial Recognition: No” but later add it, they’d need to update the sign or be in violation). In our connected era, informed consent shouldn’t be a casualty of innovation. A camera-ingredients label is a simple proposal that shines a light on hidden surveillance practices, making the invisible visible. 

By clearly labeling the capabilities of cameras, we empower the public to see through the lens of transparency – ensuring that as cameras get smarter, our policies get smarter too, keeping ethics and consent in focus even as the cameras focus on us.

Sources:

Reuters – Rite Aid quietly deployed facial recognition in 200 stores over 8 years, often without clear public notice. https://www.reuters.com/investigates/special-report/usa-riteaid-software/ 

Cartelli Videosorveglianza (EDPB guidelines summary) – CCTV signage should include any information that could surprise data subjects (e.g. third-party access, retention period). https://cartellivideosorveglianza.com/cctv-signs/cctv-surveillance-sign-gdpr-edpb-guidelines-template-smartglance/ 

CyLab/CMU – Survey: 92% of people think it’s important to inform consumers when personal data is collected.  https://www.cylab.cmu.edu/news/2020/05/27-iot-labels-consumers.html 

Pavion (Ethics of AI Surveillance) – Emphasizes transparency: individuals should know they are monitored and how their data is used. https://pavion.com/resource/balancing-ethics-and-privacy-ethical-considerations-and-privacy-concerns-for-ai-video-surveillance-in-retail/ 

CyLab/CMU – Researchers developed a prototype “nutrition label” for IoT privacy/security, allowing users to compare device data practices like nutrition facts. https://iotsecurityprivacy.org/labels/