I downloaded my Facebook Data and it is terrifying what they have.
“About You” folder
My Address book (Contact information you've added for friends and other people) - they have information on all these people in my contacts list, even the ones I’ve deleted, and the ones that don’t have facebook
I tried turning off the “Contacts Uploading” setting on facebook but it said that I had already turned that off at some point. I did manage to turn off the “Upload contacts” setting in Messenger though
I later found the “delete all contacts” option on Messenger which deleted all the imported contacts. I found the option on Facebook as well
I tried turning off the “Call and Text History Uploading” setting on both Facebook and Messenger, but the links directed me to blank pages...
‘Ads’ folder
Shows the ads that “I’m interested in” based of facebook activity and other actions
I removed some ads, but there was no remove all option, so I’ll have to manually unlike them all at some point
Shows the “advertisers who run ads using a contact list they uploaded that includes contact info you shared with them or with one of their data partners”
Now this really creeped me out
Shows the advertisers whose ads you’ve clicked on.
‘Apps and Websites’
Shows all the apps and websites that I’ve used facebook to log into, and also the date and time that I did so. This included the ones that I had “removed” - which Facebook ‘kindly’ kept in the “Removed” tab for me...
And these are just the first 3 folders!! I’m going to be going through them very thoroughly over the break and updating my privacy settings and other settings.
In the week 9 lectures, Richard pulled up this article about CIA agents being killed because of improving face recognition software. Since I did my Something Awesome Project on biometric authentication, and looked at face recognition, I decided to look at it.
My thoughts:
“Now even small third world countries have access to scanners and databases. It’s killing us. It’s awful”
And countries can share and have shared this data without people’s explicit consent, in an effort to target “crime/terrorism”
“False documentation” is not enough anymore because many airports use iris scanners in addition to fingerprint scanners and face recognition technology
You can’t really re-enrol your iris data or your fingerprint data because the system will detect that it’s already there. It’s also harder to pretend to be someone else because of this technology
“A few years back the entire CIA network in China was rolled up and there were all executed, often in a pretty brutal manner”
Because many countries have this shared database of sorts, and facial recognition technology is improving, searching for one face doesn’t take that long at all
You can’t control how other countries use the data that you have freely given them, so perhaps the only way forward is to train people young in another country, so that they can become “insiders” when they grow up.
In the week 9 tutorial, Hayden taught us about the 0-knowledge protocol, which is the method that one party can use to prove to another party that they know something, without giving away any information about that thing.
In the demonstration that we did in class (which I participated in), Hayden ‘proved’ that he was not colour blind by accurately stating whether I had swapped 2 remotes or not. The more times we performed the demonstration, the more confident I was that he was not colour blind, even though I didn’t actually know if he was telling the truth or not about the colour of the remotes.
Using the zero-knowledge protocol, how could we prove that we could steal something from any house in Kensington?
I’m thinking something like this:
Have a person stand opposite a house of their choice and ask us to exit from that house. They then turn around so that we don’t reveal how we entered the house. We then text them so they know when to turn around and see us exit the house.
This way, the other person doesn’t know how we entered the house, or if we stole anything, but they are able to see that we could exit from any house.
Edit: this is not strictly zero-knowledge, as it might be considered leaking knowledge to have them see you exiting the house. So I will keep thinking about this problem.
You are at uni, it is today. Several phones go off simultaneously, and when you check yours you see the following message:
BALLISTIC MISSILE THREAT INBOUND TO NSW AND SURROUNDING AREAS.
SEEK IMMEDIATE SHELTER.
THIS IS NOT A DRILL.
We had to pretend we were: ourselves, the uni, or the prime minister, and were asked:
What would be your immediate reaction? What would you do?
If you had an extra hour, what would you do?
If you had an extra day, what would you do?
What would you have done to be prepared for this situation?
My group searched up how long it would take a missile to travel from China or South Korea to Australia, and it would take less than 30 hour. So the time to do something became quite limited.
Yourself
Immediate response: we’ve been hacked, or there’s been an user error. We might panic after looking at surroundings and seeing other people’s responses (i.e. seek confirmation that this is real). We would then try and find out where to evacuate to and evacuate.
An extra hour doesn’t really help much since we can’t escape Sydney in an hour (if all the roads are blocked/damaged)
If we had an extra day we would drive to another state (we assumed the roads broke because of the mass fleeing, and that the roads we be ok the previous day). We might also organise a go bag or organise bunkers
To prepare, we would
Keep up with current affairs, so we know if there’s a high change of a missile attack occurring at some point in the future
Do uni by distance rather than come in every week
Learn where the tunnels/safer areas at uni are
Have go bags
Uni
We would sound the alarms, then also send a message to each person’s phone. We would also tell everyone to seek shelter
We didn’t consider the human element where high executives of the uni might evacuate first and get their friends and family to safety before considering us
We can’t do much with an extra hour
If we had an extra day, we would tell everyone not to come to uni, and to leave Sydney. We would also try and protect our assets (students, money - also dead students can’t pay money, data, machinery/equipment, research papers etc.)
For preparation, we would
Build shelters so that if we couldn’t leave in time, the students would have a safe area to evacuate to
Have evacuation drills specific to missile threats
Perhaps only have distance studies?
Prime Minister
Our group felt that there was probably some protocol in situations like this, where the prime minister would get bundled to safety regardless of how we reacted. (We felt the ACT was close enough to Sydney that the prime minister might be in danger)
Again, we failed to consider the human element where the prime minister might flee for safety without any regard for the wellbeing of the people
If we had an extra hour we would
Invoke military to try and shoot down missiles, and/or contact other nations i.e. allies to help shoot them down
We would negotiate with the country to not send the missiles
We would try to evacuate people calmly
We didn’t consider the option of surrendering
With an extra day, we would
Again, keep negotiating or organise the evacuation plan
Or surrender...
Preparation
Arms race - build a missile ourselves and have mutually assured destruction
Prepare some protocols for situations like these so people can carry them out effectively without wasting time panicking and wondering what t do
Build bunkers for the people
Make planes carry supplies - food, water, first aid etc.
I noticed that a few weeks ago with the cyber war case study, I voted for peace with Russia, but this week I went straight to arms race.
Richard wasn’t here on Tuesday, so we had a guest lecturer speak about privacy (or the lack of privacy in the modern age). Some interesting things I learnt:
Chatham House Rules - source of information shouldn’t be disclosed/revealed
All anonymised data runs the risk of being reversed. In fact, many scientists have the job of de-anonymising data
All anonymising strategies (redaction, encryption/hash, pseudonyms, statistical noise/”binning”, and aggregation” are breakable
The tools to download all your data weren’t made for users but for the government, but we can take advantage of them
Some companies are under legal obligation to you the information they have about you, but they don’t give the inferences they’ve made, only the raw data
Websites are increasingly using fingerprinting (information about the computer that you’re using e.g. screen resolution) to identify users, esp. for people that disable cookies
Evening Lecture
For the evening lecture, we had an extended seminar on rootkits, and then some lightning talks (one by Hayden on typosquatting) and also some final exam information.
Rootkits
Can provide root access without being detected. Not malicious by itself but can allow malware to be installed
There are few different types of rootkits that different depending on which level they operate on e.g. kernel level, usermode level etc.
TypoSquatting
Where a user misspells the domain name of a legitimate website, and is directed to a website that is owned/controlled by an attacker
A lot of companies buy the typosquatting domains e.g. gooogle.com redirects to the normal google
Browser Fingerprinting
Say ‘no’ to clipboard permissions otherwise websites will be able to scrap everything on your clipboard every time you enter their website
Our week 9 case study was about guns and gun shooting in America. We were asked to think about two problems:
School shootings
President assassinations
My group looked at how we could prevent and respond to school shootings.
Prevention:
Bump stocks should be made illegal (they make semi-automatic rifles act like automatic ones which are illegal in the US)
Put guns and bullets in separate safe vaults so that kids can’t steal them and use it
We were told in the discussion that this was already implemented in America, but that no one followed it because it was a hassle. Guns are part of the second amendment for the purposes of defense, and people can’t defend themselves against gun-carrying attackers with the gun dismantled in separate safes.
Have psychologists at schools that can monitor children and their emotional state, to prevent them from wanting to kill their peers. This might help deal with children that have been bullied, or express signs that their home life is not too good
Have counsellors at schools so that children can talk to them - this might not achieve anything because people often don’t want to be truthful about their emotions, and it might make some people more likely to rebel against ‘authority figures’ and became a mass murderer
Response:
Automatic shutters for blinds
Compulsory metal detectors and searchers - the second the detector is triggered, give everyone a limited time to evacuate to the nearest building and then automatically lock the doors
Bulletproof locks
Drills regarding gun safety, and repeated practice, especially when the students are outside
Have metal detectors face upwards on the boundaries of the school so we can detect if a gun has been thrown over
We determined that it was not the school’s responsibility during out of school hours (i.e. before and after school).
We also thought that the only way to really prevent gun violence was to ban guns (not possible because it’s a constitutional right to carry guns) or to have a culture change, but that’s also quite hard to accomplish right now. There’s also a lot of things that we can’t really do, because it would be too restrictive and would most likely cause people to revolt.
We wrapped up the learning content for the course this week.
Morning Lecture
We thought back to the question of what it means to be secure in an engineering sense. Some key points were:
We don’t start from scratch - we build upon what already works, what has been tried and tested to be good.
Engineers have a culture of excellence - we take pride in what we do and attempt to always produce the best results
There needs to be constant review - review of the process, the testing, the building etc
We need to pay attention to the crucial details, rather on the minute. The example given was the 350 alarms that went off at an oil rig, where only 8 or so were crucial
Evening Lecture
The two talks given in the evening lecture was on Reversing, and Cracking. I was really excited when I was able to understand a lot of the content they presented, largely due to the optional Reverse Engineering seminar/demonstration in Week 4 run by SecSoc.
This course has definitely motivated me to start CTFs and learning how to reverse/crack programs. The plan is to start learning and doing some puzzles in the holidays.
After explaining to my friend how easy it is to crack his iPhone TouchID (part of my Something Awesome Research), I have successfully convinced him to get rid of it and use an alphanumeric password instead.
When he first started it was very annoying because he had to type a long password in, but hopefully it makes his phone a little more secure, since he had cards stored on his phone that used to be ‘authenticated’ with TouchID. Now the next step is to make sure he didn’t use anything too personal in his password!
I am still in the process of convincing another friend that she needs to set up a password on her phone. I have high hopes.
The invisible man can talk to the alien and vice versa
The major can talk to the invisible man, but not vice versa
The invisible man was the major’s cadet (so we know everything about him within reason)
We’re on earth right now (and not the alien planet)
The questions we had to ask ourselves and answer were:
Is the alien A to be trusted?
Is there anything urgent we should do based on the information you obtained in your trip to the alien planet?
My group decided that the alien could not be trusted since there was a conflict of interest. It also seemed like our group was the only one that thought that the invisible man couldn’t be trusted, and that we had to establish his identity before attempting to get his information. Hayden mentioned later in the group discussion that he could have been tortured for example, and that any information the invisible man has is compromised if he was alone on the alien planet.
The information we wanted to know was
Are there low-intelligence or high-intelligence lifeforms?
Are the aliens dangerous?
Is the planet habitable?
What happened to you? (Why are you invisible/a ghost)
If the invisible man was actually there, and if we think he is who we think he is (authentication)
If the invisible’s values have stayed the same or not (decide if we can trust his message is for our benefit)
Given that the invisible man was the major’s cadet, we decided to use past history as a sort of key to authenticate him e.g. What did we do in the Battle of R with X, Y, Z. After authenticating him and his values, we could then start to ask him questions about the alien planet e.g. Is the planet’s weather like battle X or battle Y?
The problem with this is that we don’t continuously authenticate him, so we don’t know if another invisible man is speaking for him during certain questions for example.
Our first presentation was on privacy. Some key takeaways were:
It’s not about if the data is compromised, but when
We can use a different browser e.g. DuckDuckGo to increase privacy (or perhaps the illusion of privacy)
“I have nothing to hide” = “I have nothing I can think of that I want to hide”
Our second presentation was on digital forensics, and the demonstration that they gave during the lectures really made me want to try as well. I might start in the break, and then keep learning and practicing in COMP6447 hopefully.
The evening lecture component consisted of Richard reading from the book “3 Mile Island”. It told of all the things that went wrong with the system that led to the partial meltdown of the nuclear reactor, and all the responses and actions that were taken by the staff there.
There were so many little things that all just added up to cause the big catastrophe, bringing home the point of latent errors - when the error happens now but the consequences are not visible until later.
In the morning lecture this week, we looked at root cause analysis and a few other types of human weakness.
Something that I’ve learnt throughout this term is that no matter how hard we try to understand what the root of the problem was, we’ll always miss some something. We tend to focus on what makes sense to us, what’s happened before or what’s easiest to fix. This was once again reiterated by Richard. We also tend to think that an event only has one significant cause, rather than many causes.
For human weaknesses, we looked at
Honesty - how people, as complex beings, are also motivated by a whole range of things
Misdirection and limited focus - it’s hard for humans to pick the things that should be focused on. We’re too easily distracted by minute details. As Richard said in the lectures, “we should be focusing on what’s logically important, but we tend to focus on what’s physiologically salient”
Similarity matching - we try to match what’s currently happening to something similar that’s happened in the past so we have a set procedure we can happily carry out. Social engineers will try to set things up to exploit this trait
Frequency gambling - when we don’t pick the most suitable pattern for the current event, but the pattern that we’ve used the most in the past just because it’s familiar to us
Group-think syndrome - when you value group membership, harmony and being valued rather than getting things right.
OAuth is an open standard for access delegation, commonly used as a way for Internet users to grant websites or applications access to their information on other websites without giving them the passwords.
It is a standard that apps can use to provide client applications with “secure delegated access”. OAuth works over HTTP and authorises devices, APIs, servers and applications with access tokens rather than credentials.
There are two versions of OAuth: OAuth 1.0a and OAuth 2.0. There is no backwards compatibility between the two, and they cannot be used together. OAuth 2.0 is more commonly used.
In the last half hour of the tutorial, a guest tutor (Adam) gave us an introduction to format strings.
It was really interesting and I learnt a lot! I was even able to answer his first question of what printf does with the arguments (which is grab the next thing off the stack).
A few things I learnt was:
%s - takes the next thing off the stack, treats it as a pointer and prints until we get to a null byte
%s5p - prints the 5th thing off the stack
xxd - prints bytes as decimals
%p - prints it as a pointer
We can use printf and %n to write what we like to memory. This is useful in attacking other programs, and making it execute our code
Our week 7 tut was a debate between if we should have more privacy, or if we should give more information to the government.
I was initially on the more privacy team, but Hayden threw a curveball and made us swap teams.
Some good points mentioned for more privacy was:
Our telephones and internet usage etc. is already being monitored - don’t want the government to have even more information about us
In other states, the police can access your opal card travel data without a warrant
Might not trust the 3rd party companies e.g. Huawei - might have a backdoor in their systems. Some companies e.g. google are already eavesdropping - how can we trust that the government’s interests won’t change?
3rd party companies might have better security than the government - having so much data in one place makes it a big target
Data retention is quite expensive - want better allocation of money
Once you give them access to one thing, they will keep on wanting. Never ending vicious cycle
Government might become corrupt and use the information to discriminate against people
They will most likely sell your data
Some good points mentioned for less privacy was:
The government can actually do something about the information that is already being collected by outside companies
They should have more access to data to catch criminals
They can use this to find more people - e.g. more than 300 lost children were found by using security cameras and face recognition
We can increase security in public areas and make those areas more safe
We want the government to be an active participant for the wellbeing of our society, not just an observer
We can use more machine learning with increased data
We can reduce bias in face recognition, allowing for better security in the future
In the end, I still think we should be not be giving the government more data about us, given that the government can be corrupted and that they could sell your data to outside companies.
The Bhopal disaster, or Bhopal gas tragedy, was a gas leak incident that occured in December 1984. It is considered to be to world’s worst industrial disaster, with over 500000 people exposed to methyl isocyanate (MIC) gas. The official immediate death toll was 2259, with about 8000 people dying within 2 weeks of the accident.
The factory was built in 1969 to produce the pesticide Sevin with MIC as an intermediate. However, after the plant was built, other manufacturers started produced it without MIC and eventually the demand for pesticides had fallen. Despite this, production continued so there was a build up of unused MIC in the factor.
There had been 2 earlier leaks in 1976 and 1982.
In late 1984, one of the three underground 68000 liter liquid MIC storage tanks lost the ability to effectively contain most of its nitrogen pressure. The inert nitrogren gas was used to allow liquid MIC to be pumped out of each tank as needed, as well as keep impurities out of the tanks. Because the tank could no longer contain the gas pressure, liquid MIC could not be pumped out. At the time of it’s failure,, the tank contained 42 tons of liquid MIC, far more than the 50% as specified in the safety regulations.
By early December 1984, most of the plant’s MIC related safety systems were malfunctioning and in poor condition.
On the night of the accident, water was believed to have entered the clogged pipe which resulted in a runaway exothermic reaction. Two different senior refinery employees believed the psi readings were instrumentation malfunction. The situation was made worse with many security parts taken away for maintenance, or deactivated at the time.
Within 2 hours, 40 tons of MIC escaped from the tank into the atmosphere.
Analysis
The team didn’t stop producing MIC because it was cheaper than the alternative options. They valued making money over safety.
They also didn’t treat employees or surrounding citizens as assets to be protected, considering that they continued with the production of MIC despite the environment being unsafe.
There was also human error is not believing the readings, or not having a proper look before disregarding the readings.