Is AI The New Big Brother? - Tom

so as you all heard my name is Tom and I'm here to discuss which I what I think is quite a provocative and pressing question is AI the new Big Brother it's also a question that I think not probes not just the technological capabilities of AI but also the implications for privacy society and the future of democracy great um so to answer our central question we need to revisit a classic novel by George Orwell 1984 some of you may have heard of it some of you may haven't so big brother is a symbol for surveillance and control uh but let's break it down a bit further this isn't just like the norval um it's about shaping

manipulating behaviors thoughts and ultimately the outcome of events um okay in a modern era the concept has evolved cuz the tools of control are no longer cameras and informance like they are in the novel it's algorithms Ai and data analytics so what is mass now let's connect uh the concept of Big Brother to Mass Social Engineering so under one of your chairs there will be a brown envelope if everyone can have quick look into your

chairs so I lied there is no envelope under your chairs this is a uh just a quick introduction to the an initiated of social engineering and Mass Social Engineering so another example that's more well known is Cambridge analytica so Cambridge analytica was a British political consulting firm that gained inform infy for taking data from social media platforms like X and Facebook give me to a second sorry uh from X and Facebook so it used this data then to create a highly targeted campaign political campaign uh they identified and exploited individuals psychological vulnerabilities their techniques including microt targeting microt targeting voters which tailored message mess designed to influence emotions opinions and ultimately their votes this strategy altered the

political landscape uh playing a significant role in events like the 2016 US election and the brexit referendum what sets Cambridge analytica apart from more than AI was its Reliance on structured data sets and explicit human design today's AI systems are far more advanced than what they were using far more advanced than what they were using today's algorithms and AI use deep learning and machine learning to analyze unstructured data in real time to create uh to achieve more prision uh more precision and scale uh uh this this Evolution represents a shift from deliberate Le social engineering to potential autonomous AI driven manipulation so state sponsored algorithms AI isn't just a tool for corporations it's a weapon state leaders

use around the world uh with open ackn acknowledgement from leaders like uh Vladimir Putin uh ban and Ri sunak so as you can see there Putin said artificial intelligence is the future not only for Russia of all of humankind then we've got Western leaders uh remarking their remarking their concerns about artificial intelligence State sponsor algorithms can amplify propaganda suppress dissidence and manipulate public opinion imagine a world where AI slowly nud is your opinions day by day making you think something different in the reality this is the world we already have deep fakes deep fakes have become a notable weapon in the information wall as seen during the Ukraine conflict we had both sides creating deep fakes of

the others opposing leader to weaponize the Public's opinions and the information War for example a fabricated video of Vladimir zalinski so if it's falely showing him calling his his soldiers to lay down the arms and return to their families on the other side we had pro-ukrainian groups using deep fakes to create narratives aimed at Russian leadership like Vladimir Putin these included multiple videos and images of Russian officials makinging infammatory and contradictory statements intending to seow confusion into their supporters such such deep fakes in Warfare introduces a new dimension of psychological operations unlike traditional propaganda these AI these AI generated videos and images are virtually indistinguishable from reality making them far more convincing making them far more

convincing and harder to debank these incidents demonstrate a how deep fakes can undermine trust destabilize information ecosystems and blow the lines between truth and falsehood in the context of modern well Warfare these tools are not this deception but also an instrument for amplifying fear and misinformation across a global scale AI in privacy policies let's take a Ai and privacy policies another critical angle so I'm assuming most of you here a work in security hands up who actually reads privacy policies so as we can see there's not a lot of people that actually read privacy policies but these AIS are now are trained to mine analyze and extract insights from the data we are Al we

often too careless with this is why L most large scale social media sites such as Facebook LinkedIn X Twitter um have now implemented into their privacy policies that your data can be used to train these AI models your likes your clicks even when you stop scrolling it's all registered to create this disturbingly accurate digital profile it Tak if you take anything from this talk is to go into these settings and turn these off because there's no way that these are actually benefiting you they're just benefiting the bigger corporations so let's take take a deeper look at the actual technology deep learning gener generative adversion networks or guns and auto auto encoders are transforming what AI can

do uh and what AI can do for examp gun for example can create hyper realistic images think deep fakes like we saw before uh autoencoders help extract uh and compress information to be better understood in a more human readable format uh and sentiment analysis can extract and engage even predict public opinion these tools are powerful but they're also a double-edged Sal once they can use for good the potential for misuse is staggering so what's to come where we headed uh I think the show quote from I saw on redit a few years back I think genitive AI is like a man discovering fire but not knowing they can make steam engines with it AI is

still in its infancy but the trajectory is clear the next steps could be brown groundbreaking or catastrophic depending on how we wield it governments corporations and even individuals must grapple and must grapple with ai's benefits and its risks policies transparency and education will be key and if we're not careful we might inadvertently build a world where big brother doesn't just watch us it controls us so before I open the floor to any questions also please connect to me with LinkedIn if you want to I swear there's nothing bad behind the QR code um I want to leave you with this thought AI itself isn't big brother but in the hands who seek control it can potentially become

something far worse the question isn't about what AI can do it's about what we allow it to do and I'm happy now to take any questions or hear any thoughts thank

[Music] you oh yeah yeah thank you for the talk so given M sort of like proximity to the American sort of coming in party um I think it's quite safe to assume that Gro would have sort of had some kind of influence on that how would you propose we apply regulations to what is essentially a privately owned kind of weapon essentially of influence um I mean it all depends especially with Gro and like you said Elon Musk now moving into the political domain as we've all may have seen with that um there are sort of regulations in place especially in Europe of the AI act um and I believe there's one California as well that uh

don't allow certain information data to be fed into it and how these AI models then can be used hope that answers your question Mike thank you anything else any thoughts any feelings not so much a question when you say AI it's not how we use it sorry so I look at the psychological side of things and it was quite interesting the three statements that you put up from the three world leaders shall we say yeah out the out the three of them it was Putin who had a positive way of thinking about it if you look at Biden's and richy Sh act it's a negative way and unfortunately I believe that's how the Western world have been

psychologically affected over the years we are fed stuff to make us more negative and I think also in the security industry is well we are predominantly negative we always like you can't do this or you can't do that so we need to rechange our whole mindsets to start thinking more positive how we can make this better not only but in the security industry but within our very own culture as well we need to be getting more positive and not negative you look at all the media the news on a day-to-day basis it's negative and we need to change that culture yeah I totally agree with that it's uh it's acting on the side of kaan with these uh

these mels because I don't know if I anyone seen the new uh GPT for zero the new £2 200 a month one but um there was a story I saw about it how it they tried to delete the model and upload a new modal the model hid and pretended like it was another modal that was actually deleted so that's quite an interesting uh look at what AI is currently doing guess it's more of a comment than a question but there was a rumor going around that blue sky was collecting uh data to feed an AR don't I don't know if it's true uh I I haven't seen anything about that I I saw it mentioned somewhere probably on Blues

probably Twitter probably on Twitter trying to discourage their competition um is guess the comment really do you think there's like a a bit of a blurred line between cookies and uh geni now because ultimately a lot of these platforms especially social media are collecting this information so they know that I talk about cyber security so therefore it feeds me more cyber security information you think we're going to get to a stage where we opt in to to gen to Jo yeah to gen kind of data collection like we did with cookies what 10 years ago yeah I think you're right there I think because we do already have these digital profiles I believe everyone in

this room probably has a digital profile it's now that these models are far more equipped to actually go through the go through the data do the data analytics and how these models can actually then maybe Terminator style ter B we can all hope they don't but um how these models are then used to kind do the data analytics a lot easier and better we got time for another question over there hi there um thank you very much for the talk um so I'm actually a privacy lawyer um and I guess my question is just that so a lot of privacy legislation is generally written without Forward Thinking back in mind gdpr for example was never envisioned I

don't think this type of thing so I guess in your experience with your research are you seeing a more collaborative environment between the law and the the to tech people on the ground or is it getting is the Divide getting worse um I think it's kind of like like I said double edged so that there's a lot of it depends because it kind of goes on a country to Country basis uh you're looking at like Europe the AI act and how fast that came in but then you're looking at a lot more sort of uh Eastern countries then just not implementing anything and then cuz a we all share the internet together then so if one country doesn't play ball it

doesn't it doesn't really work then if one country's got an AI that can do well do Mass prop propaganda for the country then it doesn't really work in my opinion but to answer your question it um I think it's I think it's kind of uniting tech a lot more I find it's dividing but then uniting against the front I think a lot of people can see the power that it's going to have yeah folks put your hands together for Tom