AI and Cybercrime: The New Frontier in Digital Forensics

Automatic transcript of the recording (English); speaker labels are automatic. Timestamps open YouTube at that moment.

0:00Sacha: Today I have the honour of hosting Alain, an expert in digital forensics and cyber security. Alain is a retired member of the Royal Canadian Mounted Police with an impressive 29 years career in policing. Alain is also a court-recognized expert in computer and mobile forensics and has supported major investigations involving internet child exploitation, commercial crimes, organized crimes and national security. Currently, Alain is a Digital Evidence Specialist at Alphafox Forensics where he continues to contribute his expertise. He has also served as the Chief Operations Officer at TCS Forensics. Alain, it's a great pleasure to have you with me today. Welcome.

0:44Guest: Thank you, Sach. Thanks for inviting me. Quite happy to be here.

0:48Sacha: Can you share a little bit about your background in law enforcement and digital forensics?

0:55Guest: Yeah, so I think I kind of lucked out in ending where I did, but with the time, timing is always helpful. So I joined the RCMP from Montreal in the 90s. They shipped me out to beautiful Surrey, British Columbia, to south of Vancouver. And, you know, the career and the police out on the West Coast is basically just, you know, responding to to calls from the public and then you the great thing about policing is you can do so many different careers it's so many different careers within within one. So I ended up you know with the internet and the computers starting in the late 90s it was just kind of good timing for me because I had an interest in technology and science and I ended up kind of going towards those fields with working against unfortunately crimes against children internet you know abuse of children online started so I kind of computer crime career and then eventually I became a forensic examiner you know computers and we didn't really have much for cell phones it wasn't really you know like today but it really is really exploded obviously you know I worked a lot on the old blackberries and all that and yeah it's been a really interesting the last ten years of my career I did forensics work and I decided to retire and start and continue to apply those skills in a business world.

2:38Sacha: So you just mentioned there right you started at the very beginning of technology can you elaborate a little bit more on how the landscape of cybercrime is evolved over over the years?

2:51Guest: Well yeah of course yeah it started when we had pagers there was no cell phones we are we you're lucky if you had a pager and you know some of some of the listeners viewers may may not even be familiar with those but you know and one thing I like to point out is cybercrime you know people talk about cybercrime it sounds very very sexy very fancy and people think oh my god this is but cybercrime really for the most part it's it's mostly frauds it's you know they're mostly the same crimes that people are doing and it's easier to make them and faster and you can reach a bigger audience with the with the cybercrime I mean it's like you you have a podcast I mean in the old days you would have maybe like a newsletter and you would send a newsletter well how many people can you reach with a newsletter you know or they used to send those scams you know the ransomware and the all those frauds they're the prince of Nigeria you know those things existed on papers letters so it's just an evolution of of regular crimes and unless you're talking about national security or attacks on you know infrastructure you know national importance infrastructure of national importance mostly it's just your old frauds you know people trying to steal money from people uh and the priority is not changed people get surprised at that sometimes like a fraud is still a fraud even though it's happening and it's called cybercrime and it's on the computer uh a fraud is still handled like like like it was before and not all crimes have the same priority in the policing world and that's by design so

4:47Sacha: and i'm curious to get your take on what are some of the most or some significant challenges that you guys face when dealing with cybercrime in in this era of of ai you know ai has been becoming prominent and it's becoming bigger and bigger and now more advanced what are some challenges challenges that you guys have to adapt to well

5:14Guest: i think for the general public the uh it's going to make it even harder to detect uh these frauds because um you know historically uh you know there will be grammatical errors let's say or an email would read kind of weird uh you know the english wasn't really perfect and uh you know i'm going to throw out there's 75 of uh cyber crimes and frauds uh they're coming from uh from russian controlled space so uh you know a lot of them come from there and so um you know the english was not great the grammar was not great but now ai can and does uh really make it more and more human-like and it's going to be harder and harder for people to detect and uh ransomware you know all the the the the malicious codes the viruses ai is basically creating all those um all those things and it's going to be really hard to uh to identify basically because policing is always a little bit behind if you will it's a catch uh mouse in the cat and mouse game so the criminals come up with something new and then the police have to kind of try to figure out how to counter it um so ai is just you know one thing i say is police can use ai as well uh and companies you know a lot of these uh a lot not all this uh these crimes uh get to the police so companies can also help themselves by using ai to maybe filter emails before they get to to the clients or to the the employees like a lot of employees are at risk of uh falling victims for these and then it affects the whole company right mm

7:02Sacha: -hmm and that's a great segue to my next question which is in in which ways if there are in which ways is the law enforcement enforcement using or leveraging ai and generative ai to uh combat cyber crime

7:19Guest: um well speaking for um you know i've been out a few years now so they may have some little secrets that they're working on but uh even before i left the uh the software basically you know it's mostly software trying to detect or help uh let's say you can um you know in the child exploitation realm let's say ai can try to identify you know victims if you see somebody's computer or you have a lot of exhibits uh the the software companies are incorporating ai into the software to make the job of identifying victims let's say faster and uh responding to those threats faster i remember i used to work on the national security team there and we're they were coming up some tools to identify let's say a picture of somebody that's holding a gun or a rifle and so that you would try to identify those but the it wasn't really good the software if i'm being honest it wasn't really great it would make a lot of mistakes but now obviously ai is really uh kicked into high gear and it is much better at uh identifying those things um also uh you know even the phone companies uh you know text messages i mean i don't know about you but i i was i still get text messages and phone calls that you know they're scams uh so i think it's not just the police the the private companies like you know telus or bell or videotron you know they they can incorporate the ai to kind of filter out uh the criminal or the bad actors before it even gets to the the users or the clients right

9:03Sacha: so you agree and and actually yeah i do also get a lot of fraud fraudulent messages like almost every day but that's that's a that's a very good point you know it kind of makes me think on on who's accountable for um who's accountable for that like is it really the companies or is it us you know because i feel like not a lot of people get uh get thought of on how to see those messages and and really deal with them and and people still still get uh get tricked uh once in a while and then it poses problem right even in the company like if you work for a company you get an email fraudulent email and then your company doesn't do a good job of integrating ai and then you actually fall for the for the trick or the fraudulent message who's responsible what's your take on that i'm curious um what do you think well

10:01Guest: i think you know you need a balanced approach basically it's the due diligence or best practices you know if a company has has nothing to protect the employees then maybe they do have some responsibility uh they do also have responsibility to uh educate their employees and and you know uh make them aware give them awareness uh uh i've heard actually from somebody that i know who uh his company would send out these phishing emails but they're fake they're fake phishing email and they want to see who who is falling for it and if somebody falls for it then they they would get some extra uh some training some uh some extra courses right so i think that's a great thing for companies to do i know actually the rcmp would leave uh you know maybe thumb drives uh you know floating around the office to see who would plug them in to to random uh you know that was the old way of uh you know uh sending viruses it's still a way but you know so i think uh employers do need to have uh some uh you know to to to be somewhat diligent i mean i think people have to be also you know uh aware uh not to fall victims you know i get calls from people who uh they've lost you know hundreds of thousands of dollars in um in cryptocurrency scams and they they just saw it on youtube they just saw a video on youtube and they contacted the the people and and they never really researched if this was a real company uh because if you did a google search on that company within two minutes you would see that other people are complaining that it's a fraud right so i think you know i i never blame victims but i think people also have to be aware uh there's no you know there's no easy money uh nobody's gonna come and make it that easy for you yeah and if it's like that you have to really pay attention uh one thing you know you talk about companies maybe you're talking about also like social media companies or social media like facebook google you know are they kind of responsible to the monitor that stuff i think yes to a degree um you know people will post what they post but what one thing that kind of upset me when i first read about it was uh like let's say you search a national bank or you know you want to contact them and then some criminal has put up an ad on google so that they show up first so now you're getting scammed and google is make money because because they they're getting advertisement like i think that's i don't know that's wrong like there should be that that should not be happening right like uh if somebody pays money and google is the biggest one there so i'm not just pointing the finger at them but you know if they're getting money from advertisement from people that are scamming people that they need to they need to put a stop to that for sure mm

12:58Sacha: -hmm yeah for sure and you mentioned two things and actually i want to go back a little bit to something that you mentioned earlier is uh you mentioned that police is starting to use some some ai models for example uh vision right you mentioned that they were trying to to see if someone was using a gun or something like that can you give me your opinion on um the ethical consideration around using ai in in law enforcement yeah

13:28Guest: there's always going to be some ethical component with that you know that the police are doing their business um i think the obvious one is let's say like security cameras you know or surveillance uh people have a concern for um you know the state or the police and monitoring the citizens different countries of different approaches i think about like the paris olympics let's say um you know they installed a lot of uh surveillance uh cameras throughout the city to make sure it was safe and you know ai is very good let's say so you may have like i'm just making up numbers here but let's say you have 2 000 cameras but you may only have 100 people that can monitor these cameras so you know you can have ai uh look at the footage and and do some determinations uh ai is very good at that you know if you have a suitcase with nobody around it ai can pick it up uh so it's very useful for that you know um uh but uh people were asking i i read an article about you know well what about after the olympics are they going to take all these cameras down right and i don't know i haven't i haven't kept up with it but i suspect that they're going to stay there i don't know uh but the public may say uh we don't want these cameras there or you know i know in england they have a lot of cameras and people are okay with it i know in china they monitor everything but people don't have a choice the government says this is it so it depends where you live and you know in canada i think again canada's kind of a middle of the road balanced approach there we um i can see that there there's some ai uh helping uh with surveillance but that's for as far as ethical concern that's probably the number one concern for people and

15:19Sacha: i want to get to know how do you see the role of private sector digital forensics in which you are right now evolving uh in relation to to law enforcement

15:32Guest: um yeah um actually before i i just want to go back a little bit to ai with the police and the tools uh years ago there was a guy in vancouver police and he came up with a mathematical solution or model called the geographic profiling so he was uh he used to be a mathematician and he became a police officer anyway he was looking at you know like a sprinkler that sprays water and then he thought well what if you could look at all the crimes where they're happening and could could you calculate like where they're coming from like where the person might live so that's called geographic profiling and when i worked in the behavioral sciences there we uh i not myself but some people were using that and so uh i was thinking about that the other day with ai ai would uh maybe even make that even better you know or come up with new ways uh to kind of like find a new way to to solve crime or help you know crime uh through through science basically so that that's uh i just think that's an interesting concept that the geographic profiling but uh to go back to your uh question which was what were you asking me

16:43Sacha: um i was asking how do you see the role of private sector uh digital forensics evolving in relations to um law enforcement i

16:52Guest: think it's uh it's it's it's happening now to a small degree i mean uh i think it's inevitable that it's going to become more like i work uh i still work like i offer my services to the police or the rcmp if if they want to use my services at the uh at the high end you know it's it's not a lot of the work that's going to be doing themselves but uh i think the the because of the volume their volume issues you know uh if you have a case and you can't wait a year for the lab to do your examination then you need to you need to work with a private sector and i think that's okay as long as all the you know obviously the police and the government wants there's the safety requirements and uh you know for big big like big events like big cyber hacks uh i think there's a lot of expertise there's there's big cyber crime companies that that can help solve these crimes they have a lot of uh talent and expertise you know um a lot of people will work uh for the police let's say and then they get a better job but that pays more somewhere else so why why completely lose that talent you can still work with them even though if they're in a different uh a different company let's say and so i think that i think it is good and it's inevitable um i've i've been kind of preaching this for a little bit there my my uh the example i like to use is nasa uh they're working with spacex to to go to the moon and i think uh you know spacex is going to go and uh get those uh astronauts from the the boeing and they're stranded on the international space station there so i think you know why not uh if nasa can work with spacex then uh you know police can work with private sector i don't think there's any problem there yeah

18:41Sacha: that's a good point and actually how how do you or how do international boundaries complicate cybercrime investigation and how can law enforcement agencies um cooperate globally

18:56Guest: that's a tough one that's uh a lot of people are trying to solve that one for a long time and it's it was mostly it started with drug trafficking you know like years ago people those boundaries and it's the same things different different rules different laws different procedures you know different levels of uh professionalism if you will like there's corruption in other countries where yeah they know there's maybe child exploitation going on but they don't really you know maybe they make they get payoffs those kind of things are very difficult right and like i said i already said that there's a lot of cyber crimes uh like initiate from russia and so i don't know maybe they they obviously don't take it seriously that there's challenges there i don't know but it's it's the truth and um you know so for let's say the canadian police to deal with something in another country they have to go through interpol interpol you know is the the middle person and then hopefully something happens on the other end but oftentimes you you don't hear about it um i think when i worked on the child exploitation unit um we were lucky in a way because it's a crime that kind of almost everybody agrees needs to be dealt with and and so we had a virtual global task force so uh you could have police officers from many different countries and kind of all join and be part of a task force and so you're all kind of on the same team and it was kind of cross border you know i could call somebody in the us and ask for assistance there right away i don't have to go through uh ottawa and then go through uh interpol and then because that will take six months just to get a phone call so uh these things uh you know for uh building some kind of teams to deal with this and obviously cooperation cooperation and you need cooperation on uh or interest in all the parties so hopefully uh you know i know that's it's always moving along and uh hopefully uh we're not going to solve all the problems today there but a little bit at a time a little bit at a time and

21:13Sacha: i know you're you're very interested in technology and you mentioned it what emerging technologies do you see or do you think will have like a big impact on on uh cybercrime but also digital forensics in the coming years or even right now um

21:29Guest: well ai is moving very quickly and i've seen you know the numbers and i think it's really hard to even understand how it it's it's growing because it's not just it's not linear linear you know it's not twice as much in a year you know it's not twice as much next year and four times a year after you know it's like it's exponential that it's learning at that rate and then the technology i mean there's a thing called quantum computers people hear about it but basically quantum computing makes things even faster and um so you've got ai that's growing fast and then you've got quantum computers you know the hardware is is growing uh fast so that's a recipe for very very very fast and so i don't think we can even understand like people we think in our lifetime but in a hundred years and a thousand you know in 500 years what's uh ai going to be i mean who knows um you know maybe it's already happened and we're living in a simulation i have to throw that in there because uh i've been reading about that you know that this is all a simulation anyway and we're just we're just reinventing what's already existing but we we can talk about that another time

22:51Sacha: but how how do you see ai impacting or like disrupting digital forensics do you see like specific use cases or specific applications where people are using or will use uh ai in the future

23:09Guest: um yeah i mean like again for myself i use uh there's ai in the software i use and um it's um you know still being developed and they're incorporating uh more ai and actually you can use it both ways like i get uh i get calls people sometimes they'll say hey can you look at this video and tell me if it's real or not or was this ai generated or this audio file you know somebody's saying this is a recording that they made of a discussion well can you tell me if it's ai generated right and you know it you can examine it it's it's not you know it's not easy i would say so you know ai tools will um will be able to help uh and and kind of sort these things out or make it easier uh but one thing you have to remember though it's like um if you're asking ai to find out if something is real or fake you're also teaching ai how to make better fakes because it's a it's a network it's a neural link if you will like basically all the ais if you say something to chat gpt you're not saying it to one person you're saying it to everybody right so then it's kind of like it's also spreading so if you use ai to fight crime you're also teaching it to make better crime so it's um you know that's going to make things a little bit interesting i i think right so like give you the example of a car so if you get in a car accident you know you might tell your family your friends oh i had this car accident and this is what happened and your friend you know they may learn from it and say oh okay yeah better you know if that situation shows up i better not do that right but if an automatic uh those automatic driving cars if they get in a car accident well all the cars that are running that software will all know about it and so that's the kind of knowledge sharing that's happening and so it's very it's instantaneous almost and it's very fast right so it's effective but yeah you have to be aware of it

25:19Sacha: yeah in especially in the area of crime it's a it's a very vicious circle as you say like because you're you're teaching the ai to become better but then the people that use ai to do crime is also getting better so it's it is a really vicious circle and it's a very interesting point do you have any uh any cases where you can that you can share with you know ai or even advanced technologies uh playing a significant role in either commenting or even solving uh like a crime a cyber crime mm

25:55Guest: -hmm uh well specific cases i don't think i have any offhand there but um um you know i've read about uh ai being used to solve like um i think for the the biggest one is being used right now is like unsolved murders um you know you've got these cold cases and you can have ai kind of do a review and and and see what what has been omit if anything's been missed you know humans have a we have a great capacity for certain things but there's also maybe we missed something that's kind of obvious ai may pick that up i know there's some cases like that where um ai was able to reconstruct uh facial features let's say of a of a skeleton and and rebuild a face so that they could come up with a actual you know a look-alike and then they were able to solve a crime that way um yeah i think um i don't know if they still do those composite drawings you know you go in and you tell the police officer uh to draw a face with a pen and pencil there i'm actually not sure if that's still happening they're probably using some kind of uh ai tools or i think they will be in the future uh that's for sure um so the that i think that's the obvious ones um you know and then criminals hey they're coming up with uh all sorts of the new tools uh you know there's there's uh specifics so what they do is they'll take like an older version of chat gpt let's say to to do some crime because chat gpt is like any software it gets patched you know it gets it's good block and they block stuff right especially like the child exploitation images like unfortunately criminals can use ai to create images of um you know children and sexual uh activities and so that's that's the downside but you know you can't do that with the current version so but they're using an older version on a standalone system right um so those are the kind of things that uh again you know it's cat and mouse they try to do this and we please try to do that but um yeah it's it's definitely going to continue maybe

28:08Sacha: it makes me think about um i don't know if you've heard about that but like a couple of months ago even maybe a little more like five months ago there was this huge case in in the u.s about uh robo calls and it was impersonating joe biden so people would get calls and it was basically the fake voice of of joe biden so it was an ai calling someone and requesting i think it was money or something but anyways it was a scam it's pretty impressive what they're uh they're able to do and it leads me to the next point which is right now there is a bit of a i wouldn't call it a war but there is like a some sort of debate between open source ai and uh closed source ai so i don't know if you're aware but some companies like meta they have they're releasing their new models open source which means everyone has access to the code and everyone can basically use that you know on a standalone computer if they have like good enough uh computers obviously but they have access to the models so they can do whatever they want with it on the opposite side you have open ai and chat gpt which as you mentioned like it's it's more closed no one has actually access to the code so they can't really replicate and and then open ai when they find a bug with something that you know it shouldn't be able to do they close it they patch it as you mentioned um so what what do you think about that like do you think it's very um tricky or um maybe even it shouldn't be legal to have open source because then you know you have countries and you know it's it's part of the debate like china they can just take the code from meta for example and then use it against the us or canada like allies what do you think about uh what's going on from a law perspective

30:13Guest: well i don't think yeah i mean i don't think you can block um open source code to it's gonna either leaked out or you know somebody's gonna steal it like it's gonna be out there anyway at some point and you know a or ai can also create its own ai you know so i think it's like uh the old saying you know you can't put the genie back in the bottle um you know some people ask you like oh well we should put limits on ai you know so so it doesn't do this or it doesn't do that but uh when i do my presentations i always mention that not all you know it's a competitive world like that's the world we whether you agree with it or not it's all about money and a company like chad gpt you know they want to make money and they're not going to limit their capabilities if uh russia or china is going to make an ai that's going to be better than theirs so you know because we have this competitive uh model and not you know until we're one planet and one government um you know these things are going to be competitive and so nobody's gonna um uh try to limit their their ai either right and then um uh the open or the open uh you know uh open source model is interesting i mean people are going to do stuff with that like it's kind of like linux and windows right i mean you got windows and apple software and and then you got uh the linux models so um people are going to argue which one is better they're gonna argue about that but i think the it's kind of the opportunities are limitless now and you can't you can't go back whatever whatever you can't block it basically you can't block the signal what

32:12Sacha: advice would you give someone um or even businesses to protect themselves from ai enable cyber crimes well

32:21Guest: for businesses um you know i would definitely recommend um uh hiring the services of uh or you know if you're a big company is going to have in-house people they're going to hire they're going to have a department to do that uh smaller companies are you know i encourage them to seek out the services of experts who can do like penetration testing let's say but you know that's been around for a while basically a company will try to hack into a company and then provide a report to see you know if they were successful what can they do better suggestions so i think that's happening um i don't offer those services myself but you know companies that are going to be doing ai penetration testing if you will to uh to try to to to uh improve the how a business is dealing with things but definitely you can't do it alone and and and expect um you know high success rate there you're gonna have to uh get the services of other people and experts and try to build the best uh model and then for the for the citizens you know the average person is uh it's just uh it's kind of like the buyer beware you know if it's if it's too good to be true uh like personally i don't even click links on the phone i don't like it like if i if i get any transfer or something send it to my email because then on a monitor on a browser with a computer monitor you can see more you know where this email is coming from and um i think it's much easier for somebody to scam somebody on a phone i don't really but the the younger generation loves their phones and uh it's hard to it's hard to get them some people don't even have computers they're just and a phone is a computer by definition but um yeah it's it's it's it's a tough one but basically if you don't expect money don't click a link or if you don't owe money like people get bills they're like panicked oh they're saying amazon says uh i owe them money for this uh no you know did you order something like if you didn't order it like don't worry about it let them let them come to you and i think uh society in general is maybe getting a little more like laissez-faire you know but maybe a laissez-faire attitude you know just let them come to you and uh let them prove that it's real don't don't be so people are very trusting by nature me as a police officer it's the other way around i'm always been suspicious and i question everything people need to develop a little bit of that skills sir but um and educate themselves and don't try you know i can't i can't uh you know even my kids they're watching tiktok and they're getting new news from tiktok and i'm like that's not news but that ship has sailed that's that's what people are going with even the even the older generation there they see something on youtube or on instagram and facebook and they they just assume that it's real um so don't believe don't believe it right away always double check

35:36Sacha: that's a very good point you know in in all of the conversations i've had with uh leaders like you in technology it came it became apparent that we have to become more uh better at critical thinking especially with the rise of ai you know as much as uh it's great at generating text images and everything so as much as it is good to augment someone the more we have to to become better at soft skills like critical thinking like communication between you know one another and and those soft skills because obviously ai and especially generative ai is becoming better at replacing hard skills like coding like you know different stuff but anyways i think that's a very good point like critical thinking in general especially with with frauds and and the stuff that ai is able to do nowadays it's becoming increasingly important um it was it was a great conversation and i want to give you the floor to say any last word to to the audience whatever you want to share

36:49Guest: um well uh again it's just kind of like being uh being aware of uh all the fake content that's coming our way and even myself you know somebody showed me a picture of uh and i i don't make political uh sides it was a picture of jd vance apparently he was in university and he's dressed as a woman and my first thought was like well that's interesting because he's kind of against that i think but then i thought about it after i don't even know if that's a real picture like where did this come from and then i went and i did some searching and i never saw really any website or any news that said like this is a real picture so unfortunately it's you know question everything like don't don't be quick to react don't be quick to judge and uh always just double check your facts you know i kind of i can be annoying to my friends and family because if they say something and i'm like well i'm gonna look that up and and so i get to be the annoying guy that way but uh you know i think the i think there's a little bit of safety in in that so that's that's kind of like my last uh question everything yeah

37:57Sacha: i fully agree sometimes annoying and being safe is uh actually the good side to be on alan thank you so much for joining me okay

38:06Guest: great thanks for inviting me you