I’ve collected a lot of privacy tips over the years, ranging from somewhat common to very niche. After seeing this post, I wanted to share them, along with ways to mitigate each of them. I’ve tried to find sources for each of them, but some of them simply don’t have good sources. You can help by finding good sources (Wikipedia, research papers) covering the specific topic. As with all my posts, this is not AI-generated, just well-structured.
Good sources
Glove Prints
Problem: Thin gloves like surgical gloves can still leave fingerprints on surfaces.
Source: https://en.wikipedia.org/wiki/Glove_prints
Mitigation: Wear thick, textured gloves
Main Hum
Problem: The hum of the electrical grid can be used to determine when, and sometimes where, an audio recording took place.
Source: https://en.wikipedia.org/wiki/Electrical_network_frequency_analysis
Mitigation: Only record audio in places far away from electrical appliances.
Stylometry
Problem: Your writing style can uniquely identify you.
Source: https://en.wikipedia.org/wiki/Stylometry
Mitigation: Use AI rewriting tools to obfuscate your writing style.
Infrared Glasses
Problem: Sunglasses don’t block infrared cameras.
Source: https://www.reflectacles.com/irlenses
Solution: Buy infrared blocking/reflecting sunglasses such as Reflectacles.
Tracking Dots
Problem: Printers add tracking dots that identify which printer was used to print a document and when it was printed.
Source: https://en.wikipedia.org/wiki/Printer_tracking_dots
Mitigation: Use a printer that may not come with this feature (such as the upcoming Open Printer or use public printers like those found in libraries.
WiFi Motion
Problem: Wi-Fi routers are able to track motion under specific conditions.
Source: https://www.cognitivesystems.com/wifi-motion/ (and others)
Mitigation: Radio jammers may provide some protection against this, since the technology is still fragile.
PrintListener
Problem: Your fingerprint can be uncovered using the sound of your finger gliding across your screen.
Source: https://www.ndss-symposium.org/wp-content/uploads/2024-618-paper.pdf
Mitigation: Use a ballpoint pen with a stylus tip, which are somewhat common to find at most events.
Laser Microphone
Problem: You can record audio using the vibration of nearby objects, like windows.
Source: https://en.wikipedia.org/wiki/Laser_microphone
Mitigation: Have sensitive conversations in closed-off or soundproof rooms or outside in remote areas.
Keystroke Recording
Problem: Your keystrokes can be uncovered using audio recordings of typing.
Source: https://arxiv.org/pdf/2403.08740
Mitigation: Use virtual keyboards.
Typing Patterns
Problem: You can be identified based on your typing patterns.
Source: https://expertbeacon.com/your-typing-style-is-as-unique-as-your-fingerprint-heres-what-that-means/ (and others)
Mitigation: Type into a text editor and copy-paste into the form you want to submit.
Unshredder
Problem: Shredded documents can be reconstructed.
Source: https://www.unshredder.com/ (and others)
Mitigation: Burn sensitive documents or use piranha solution.
Microbial Cloud
Problem: You can be identified using your unique microbial cloud.
Source: https://peerj.com/articles/1258/
Mitigation: Shower obsessively, I’m not sure.
Gait Recognition
Problem: The way you walk can identify you.
Source: https://en.wikipedia.org/wiki/Gait_analysis
Mitigation: Place gravel in your shoes or intentionally walk funny.
Store Tracking
Problem: Stores track your location using Wi-Fi and Bluetooth signals from your device.
Source: https://en.wikipedia.org/wiki/Indoor_positioning_system (and others)
Mitigation: Disable Wi-Fi and Bluetooth when not in use.
Gyrophone
Problem: Gyroscopes can record audio
Source: https://crypto.stanford.edu/gyrophone/files/gyromic.pdf
Mitigation: Disable sensor permissions for apps that don’t need it on GrapheneOS.
Camera Styles
Problem: Cameras can be identified using picture styles.
Source: https://www.scientificamerican.com/article/tracing-photos-back-to-the-camera-that-snapped-them/
Mitigation: Use Polaroid cameras or obfuscate pictures before sharing them.
Amazon Sidewalk
Problem: Offline Amazon devices can access the internet by communicating with nearby online Amazon devices.
Source: https://en.wikipedia.org/wiki/Amazon_Sidewalk
Mitigation: Don’t use Amazon devices.
Deep-TEMPEST
Problem: Data sent over a wired connection (like HDMI) can be received wirelessly.
Source: https://arxiv.org/pdf/2407.09717
Mitigation: Use shielded cables or encrypt wired connections in transit.
Cartridge Memory
Problem: Offline printers exfiltrate data through ink cartridges.
Source: https://support.hp.com/us-en/document/ish_6681254-6681301-16
Mitigation: Use a printer that does not come with this feature (such as the upcoming Open Printer.
WhoFi
Problem: You can be identified based on how your body blocks Wi-Fi signals.
Source: https://www.theregister.com/2025/07/22/whofi_wifi_identifier/
Mitigation: Use a Radio jammer.
Stingray
Problem: Most phones are vulnerable to Stingray attacks.
Source: https://en.wikipedia.org/wiki/Stingray_phone_tracker
Mitigation: Use Rayhunter to detect Stingray attacks.
EICAR
Problem: Surveillance cameras can scan QR codes.
Source: https://www.linkedin.com/pulse/qr-code-car-hack-fun-impractical-w-garrett-myler
Mitigation: Some surveillance cameras will crash if they scan a QR code with an EICAR Anti-Virus Test File.
Pulse-Fi
Problem: Wi-Fi can measure your heartrate
Source: https://ieeexplore.ieee.org/abstract/document/11096342
Mitigation: Cover your skin to reduce effectiveness.
Mediocre sources
Security Envelopes
Problem: The contents of envelopes can be read without being opened by seeing through the paper.
Source: General knowledge.
Mitigation: Use security envelopes or thick packaging.
Branding
Problem: Even when fully covered, the clothing you wear can identify you.
Source: General knowledge.
Mitigation: Use plain unbranded clothing bought secondhand from yard sales or Goodwill.
Body Shape
Problem: Even when fully covered, your body shape can identify you.
Source: General knowledge.
Mitigation: Wear baggy black clothing like down jackets.
ID Address
Problem: State ID cards have your address listed on them.
Source: General knowledge.
Mitigation: Passports are valid photo IDs that do not have your address listed.
Bluetooth Contacts
Problem: Any Bluetooth device you pair your phone to can access your contacts.
Source: Various online discussions.
Mitigation: Use a phone that allows you to deny that permission when pairing Bluetooth devices.
Voice Changers
Problem: Voice changers can be reversed.
Source: Various online discussions.
Mitigation: Use voice changers that highly obfuscate your voice (beyond something like a pitch shift) or use a program that converts live speech-to-text into live text-to-speech.
Analog Speakers
Problem: Speakers can be used as microphones.
Source: Various online discussions. You can try this yourself by plugging a speaker into your microphone jack.
Mitigation: Use built-in speakers or remove all speakers entirely.
Spray Paint
Problem: Spray paint can’t blackout surveillance camera lenses.
Source: Various online videos of failed criminals.
Mitigation: Use duct tape or other adhesives that are not only cheaper but provide better protection.
Censorship
Problem: Pixelated or blurred images can be reversed.
Source: Various reversal tools.
Mitigation: Censor information using solid boxes.
Bad sources
Radio
Problem: Receiving antennas can be geolocated using signal interference.
Source: I couldn’t find a suitable source covering this specifically. This is a general source.
Mitigation: Receive media via offline methods.
Vehicle Telemetry
Problem: Cars will collect telemetry locally to be shared when you visit auto repair shops.
Source: I couldn’t find a suitable source.
Mitigation: Use dumb cars or commute using bicycles, buses, etc.
Detaining
Problem: Police can detain you without a valid reason if you carry an ID.
Source: I couldn’t find a suitable source.
Mitigation: Don’t carry your ID unless you need to.
Copper Strips
Problem: Buildings can be wiretapped using microphones as thin as paint.
Source: I couldn’t find a suitable source.
Mitigation: Have sensitive conversations in areas unlikely to be wiretapped, or buildings with plain concrete walls.
Biometric Unlock
Problem: Police don’t require a warrant to unlock your phone using biometrics.
Source: Various sources have conflicting information.
Mitigation: Lock your phone without biometrics or use two-factor fingerprint unlock on GrapheneOS.
Pedometers
Problem: Pedometers can reveal information such as your religion.
Source: A website I still can’t find.
Mitigation: Use dumb pedometers that don’t include timestamps.
I mean if any of these are a threat for you. Your screwed anyways
I’ve never heard of infrared glasses, this is probably the most interesting/useful thing on this list for me. Too bad Reflectacles is an American company though. Anyone have any experience with them? I’m willing to give money to US companies under certain circumstances if they do good things, like Signal. If Reflectacles does good work, I’m not opposed to placing an order.
If your threat model involves all these, then you can only be one person, and he has already been arrested due to stylistics. /joking
BTW your advice to use AI against writing style fingerprinting is not what I have heard, and some people don’t want to use AI, especially OpenAI. You should at least make your remedy about local models, but those are not as good as the commercial ones.
The correct response here is: style guides.
Style guides are specifically designed to make multiple staff writers to all sound the same. There are tools like back-and-forth translation and reading level analyzers that you can use offline to minimize peculiarities in your writing.
But is is all very cumbersome and error prone, and for low threat scenarios just mimicking another person at a lower reading level than yourself is the most accessible method.
Pedometers
I thought they’re used to catch pedos.
To measure them, actually!
Some of these might be a bit extreme in my application, but they are certainly well worth keeping in mind.
Problem: Shredded documents can be reconstructed.
Reminds me of this:
During the Iran hostage crisis in 1979, Iranian students who seized the U.S. embassy in Tehran spent years reconstructing shredded documents left behind by American officials. This effort allowed them to recover sensitive intelligence and operational accounts, which were later published and sold in Tehran.
It’s worth noting that the process took several years to accomplish.
Problem: Pixelated or blurred images can be reversed. Mitigation: Censor information using solid boxes.
I’m no expert, but if pictures can be ‘unswireled’ a la Christopher Paul Neil the Mr. Swirl Face, could the black box be lifted revealing the contents of what it covered.?
could the black box be lifted revealing the contents of what it covered
If we’re talking raster images, then no, you’re not using some mathematical model to morph an image, you’re just overwriting pixels arbitrarily, so there would be no way to recover what’s “under” the box because there’s no second layer.
If instead it’s something like a PDF (vector), then placing an object over another without flattening the result would allow you to move it out of the way to reveal what’s under.I would add, of course, make sure that the black box covers enough to prevent the possibility of inferring what’s under, so if it’s text or other organic things that can be “statistically” regenerated with some model, don’t leave any borders that can identify the item, reconstructed data may be synthetic, but can provide informed guesses
This is not generally useful.
Including legally questionable techniques like fingerprint and handwriting recognition as things to be worried about and not including the effects your recommended mitigation might have (one fantastic example is the reply about how a signal jammer will make you more suspicious and not less!) makes this seem like more of a post meant to confuse and sow fear as opposed to a post intended to help people make good decisions.
The example that stood out to me most (they almost all made me raise an eyebrow but this one was beyond the pale) is where you suggest mitigating the threat of your anonymity being breached by not carrying Id to styme cops.
When you don’t have id and the cops take you in, assuming you don’t identify yourself on a phone call out which is recorded, your lawyer doesn’t identify you or you don’t just crack under pretrial detention, you will be fingerprinted, photographed and searched against a wide ranging series of databases including id photos made for third party verification and private security camera records.
Which makes it much more likely that fingerprint or handwriting evidence against you can be admissible because at that point you’re “known to law enforcement”, “in their system” and your “handwriting was on file”.
This post falls broadly into the category of white noise at best.
i am reminded of the Linux.Conf.Au 2017 seminar “how to disappear” (iirc) which lists all the then existant ways of tripping up surveillance (physical device • AI processing • identifyability of your personhood) involving atleast reflex-trinkets to disturb surveillance cameras especially at night, secondly wearing stickers that trip up AI models and inducing hallucination as well as shining blue light on points of your face to appear as a different person entirely to the AI model, thirdly either covering specifically the nose & eyes since these contained the most identifiable points in a face (according to seminar speaker’s presentation.) just recollected off the top of my head.
also dot-matrix writers to ward off printer software.
also also remined by my converting image files to plaintext formats like .xpm .xpg .xpb .ppm .ppg .ppb et al. that strips away all meta-data in the image file from my testing. then altering it. then converting that altered plaintext image file to whatever desired fileformat.
as for recordings of the space you’re in: i recall that noise-jammer that obfuscates AI models fed the recordings. just a background noise to play. i forget the channel that presented this, professional musician with a company that also develops anti-AI treatment to audio files like nightshade or glaze but actually unstrippable. he showcased the anti-AI audio jammer.
as for actual listening in… blankets covering both of your faces, loud music in background that plays at the same tones you’re speaking in, then your company whispers?
theoretically: do clothes with metal fibres woven in (faraday cage style) hinder Pulse-Fi & WhoFi ? is there a way to trip up detection by appearing as many heartbeats & a mass of people? that same conference talk brought up textile pattern with facial prints that get recognized by AI models as additional people.
truly appreciate the post
also time to put gradients on images with text like it’s a pre-Y2K slide
Great write up.
As an amateur radio enthusiast, I’d like to add a bit of info to the radio related tips:
Wifi Motion & WhoFi
Depending on how it works, a radio jammer will almost definetly make you more identifiable and not less. Think of it a bit like being in a dark room. You can sorta see other people (or their silhouettes), but if someone turns on a torch, then you can definetly see the torch. (And any jammer with enough power to “blind” a wifi router at any more than a few meters away only works at 2.4 GHz, or is probably illegal.)
Radio
(or “locating recievers”) This is physically impossible.
Now locate a radio transmitter is easy. Ham radio clubs often do regular “fox hunts” where teams try to find a hidden transmitter as fast as possible.
However, most digital devices are transmitting and recieving at the same time, for wifi or bluetooth. Even if you’re just downloading something, you are still transmitting (I think that’s probably what you meant).
Another note on Wifi and Bluetooth
This might be common knowledge here, I think I’ve seen a post about it before. When wifi/bluetooth are on, your phone actively searches for nearby networks/devices to connect to. It specifically looks for known networks/devices, and fingerprinting your phone based on which networks/devices it knows is incredibly easy.
So, when you aren’t using it, turn wifi/bluetooth off.
Also a note on Gait Recognition
Walking funny usually doesn’t work.
(sorry, no sources here, I’m in a bit of a hurry)
Think of it a bit like being in a dark room. You can sorta see other people (or their silhouettes), but if someone turns on a torch, then you can definetly see the torch.
IIRC recent studies show that this method can identify individuals with higher specificity than you describe here. The OP didn’t specify threat models, but provided general privacy advice. Moving around town with a jammer is a physical parallel of fingerprinting an anonymous browser (It’s this mysterious user again).
But if your threat model requires you are not placed in a specific place at a specific time, then just having the jammer on in this place will not identify you on its own. Then it also depends on how many people are also using jammers. If only spies used Tor, it would be very easy to smoke them out, but the rest of Tor users serve as decoy for the spies.
So the dark room analogy is not a good fit here, and it is potentially dangerous for people under certain threat models. Just setting the record gay, with all due respect.
Walking funny usually doesn’t work.
The portable cone of silence?
Very nice write up! I didn’t know a few things on here
damn this so good
