Welcome to the KuppingerCole Analyst Chat. This is episode 201. I'm your host. My name is Matthias Reinwarth I'm an analyst and an advisor with KuppingerCole Analysts. My guest today, so the first real guest for 2024, is Marina Iantorno She is a research analyst with KuppingerCole. Hi Marina.
Hi Matthias how are you? It is very nice to be here and to be the first one of the year.
Absolutely, it's great to have you. And we want to open that year, we want to open it up with the topics that we will see in the upcoming EIC conference. And we want to cover topics that are just topical. So these are all around, everybody's talking about that. And we want to talk today, again, we did that before, but we do want to do it from a different angle about the future of cybersecurity or about the intersection of AI, machine learning and cybersecurity. So if we talk about that topic, what would you consider to be the main risks when it's coming to the rise of the AI in terms of cybersecurity and related to cybersecurity?
Well, what we need to think is, of course, having technology is great, but new technologies also carry risk. And this is what is happening with artificial intelligence. Starting with automation, for example, of course, automation is great. It reduces a lot of time. But at the same time, we have to remember that this is a machine, so that we are training. So then AI not always can make the correct judgment or the correct decision. So then eventually it could happen that there could be a mistake and that mistake could affect companies in the unfortunate event of a cyber attack. On the other hand, what is happening with data security and data privacy because we need this world to understand that artificial intelligence and machine learning, they require large volumes of data. And sometimes this data could be sensitive and it could raise concerns in organizations regarding data breach. So what happens if there is a data breach with all this data that we have used to train the model? And another risk that I would say that this one is the most common one, is the over-reliance on artificial intelligence. Artificial intelligence is great. Don't take me wrong. I am a huge fan of AI. This is my field. But we need to understand that it still needs human monitoring. And sometimes what happens is that some users, they just rely entirely on AI. And there are some attacks that are very complex. And it is important to understand that, of course, AI is great. It can help. But with the sophistication that it is changing with the time, the cyber attacks or the cyber threats, the system will need to be trained as well with the time. So then it is important to still have human monitoring. So then the overreliance could be a big deal.
Right. And you've mentioned that already as part of your answer. There's also the risk of AI generated cyber threats. So the AI put in the hands of the evil people. So what can defenders then do to protect their organization against these AI generated threats, because they will adapt as well. And they don't have a problem with things going wrong too much.
Of course. So the main thing here is trying to understand that AI is a help. Okay. So it's a tool that will help organizations. Now the best way to protect the company is having a comprehensive security strategy. And these security strategies should combine traditional cybersecurity measures and AI models. So what does it mean? It means that we still need to have, for example, a cybersecurity awareness training. Well, the people who work in the organization should actually understand what is happening. And I believe that the training that employees in organizations would have would help them to actually engage to be more informed about what is happening. And nowadays with the new AI generated threats and it would help organizations in the defense. Now, on the other hand, using AI to train models, for example, or using machine learning to detect anomalies, it is actually very good because it would help organizations to actually reduce the time in identifying risk and mitigate it. So then a combination of both would be fantastic. This is what I would say.
Okay, when it comes to adapting to these new technologies, be it in defense and also being prepared for the cyber threats that they are changing, what would be a right strategy for an organization to move forward? They would just not lay off their existing cybersecurity team and throw away the existing solutions and replace it with AI. What would be a good way of integrating all of this into a common... consolidated cybersecurity strategy.
Well, of course, the idea is not just replace everything that is in the organization, but there is not only one answer here because every organization will have different risks and different needs and different profiles. So it is not the same if we talk, for example, about a law firm than a warehouse, okay, because they manage different kinds of information. Now, a strategy that could include, for example, AI anomaly detection system using machine learning, for example, to identify unusual behaviors or unusual movement. Machine learning is great, for example, nowadays with the emails. So we receive so many phishing emails, I would say daily, and every organization probably does. So then the machine learning models are great to actually filter them because they can detect phishing attempts um, more often rather than less times. And, um, also having as well AI generated vulnerability assessments. So it is important to actually integrate AI solutions that offer transparency and explainability. And, uh, I remember we, we had another episode where we talked a little bit about explainability and in this sense for cybersecurity, particularly it is very important because the cybersecurity professionals should be able to understand what the AI system is doing and why it is important to actually train the model. And if the cybersecurity professionals understand what is happening behind, then it would be easier for the team to trust the model. And of course, if we want to switch towards the use of more artificial intelligence and cybersecurity, this will be necessary at a certain point. Of course, monitoring, it is vital, it is very important, but we need to understand that AI, it's a great tool. So all the AI systems that we have nowadays are actually helping to reduce the time, to reduce the resources needed. And also in every digital industry, it is touching all the spheres and it will also affect cybersecurity.
Yeah, and I think you're absolutely right, because we just covered the topics that you hear everywhere. So when it comes to cybersecurity and AI, everybody's talking about the evil attackers that leverage AI for better, more profound attacks. And on the other hand, we say, hey, let's use AI also for protection, for defense. So this is very limited to just this attacker versus defender stuff. But that cannot be all. Are there other ways, maybe more efficient ways also to leverage AI when it comes to cybersecurity?
Absolutely. AI has the ability to transform cybersecurity above and beyond. So it is not only about attack or defense. We need to understand as well that AI can actually help to automate the reaction, for example, to certain cyber incidents. It could help to mitigate the risk. It could help with the time detection. And it is also very good to actually prepare training for people. So we mentioned just a couple of minutes ago about the security awareness training. And there are certain companies that they are using already generative AI to create different scenarios. So then it would actually allow organization to simulate different cyber attacks. And these cyber attacks that could be human generated or even AI generated. So then these scenarios would be realistic. And people will actually be up to date, let's say, with this situation. And especially, they will know what is happening in the real world. And they will know how to react. And it is not only for cybersecurity professionals. It goes in the entire chain of the organization. And I believe that many companies will switch towards that.
As I said in the beginning of this episode, we are moving towards EIC, our big European Identity and Cloud Conference, and we are currently working on the topics that we will cover there. So one important aspect that we will also cover in that podcast again and again are trends and predictions, emerging trends, in my case, in the IAM sphere, in your case, in the cybersecurity slash AI sphere. So what do you expect to be important topics when we talk about AI and cybersecurity in the next months, in the next half year? What are emerging trends?
Well, there are several progresses in AI. So I would say that it is evolving super fast. One of the main points that I could actually mention is about encryption, for example. Some of the encryption methods could be already obsolete. And AI is offering new ways to transform the data. One of them is synthetic data. It is something that not many people was talking about before, but nowadays, it is super important. Because in the end, the synthetic data, it is fake data that masks the real inputs. And not only that, so companies also can generate more data, maintaining the same features. And therefore, it would be very, very helpful to train machine learning models and to use AI to train the system and to make it better with the time. Because as more data that you use, if it is a good quality, then better will be the result. Now, on the other hand, we can actually mention as well the centralized security solutions. So the integration between the blockchain technology could offer decentralized security. And at the same time, it would actually make identity management and data security more robust. And AI is involved in all this. Now, this year, the European Commission, they created the EU AI Act. And this is a legislation that it is actually ensuring or intends to regulate the use of artificial intelligence in the European region. I believe that it will happen in other regions as well. Now it is happening in Europe. You know that here the regulations are very strict. And every year they are becoming stricter. So then this regulation is ensuring safety, transparency, and respecting the rights and freedom of the individuals in Europe. So believe it or not, with all these progresses with AI, it is important to consider as well the ethics, what is happening behind, what are the limits, and how to protect, well, what we mentioned just here, the rights and the freedom of the individuals if their data is used, for example. And it will eventually affect cybersecurity and identity access management. If you think about this, for example, in the cross border. So now there are virtual IDs, virtual wallets. So all this information that you are using in a country and moving to another one is current information. And the point is what are the governments or organizations doing with this data? How is it handled? So then it is very important to have regulations that actually help as the EU AI act. And there are many changes coming up. So then I believe that it is ongoing. That's why it would be great to see different speakers at the EIC, we will have a special track every day dedicated purely to artificial intelligence.
Right, and it is exactly the right time right now to reach out to the team, to you Marina, to Jörg who does the agenda planning process just right now. I'm looking forward to EIC. I will work towards EIC. It will be in June this year, not in May as usual, but it will be in Berlin again. It will be at Alexanderplatz. It will be the... place to be when it comes to identity and access management and much more beyond identity related security and of course AI. Thanks again Marina for being my guest today for highlighting this intersection of AI and cybersecurity before we close down, any final thoughts that you want to contribute or just close down?
Thank you so much, Matthias. It was great being here and I totally agree with you. It would be great to have people who want to increase their networking because it is a great place to actually meet professionals. You know, we have great speakers every year and I am very, very excited about this new upgrade reality track that will be absolutely AI related. So I'm looking forward to meeting you there.
Absolutely. Thank you very much again, Marina. Looking forward to having you soon again. I think we will talk again before EIC and do more all things AI. And until then... yeah, absolutely. And until then, thank you very much and goodbye to the audience and goodbye to you, Marina. Thank you.
I hope so. Thank you. Bye-bye.
