The AI Dilemma: Protecting Ourselves from the Dark Side of Tech
Artificial intelligence (AI) reveals its versatility uniquely in the hands of different people. Evgeny Chereshev, founder and CEO of MTS Innovation Center, recently elaborated on this during a Smartpress event. For some, AI can manifest as a digital avatar, allowing a user to communicate in foreign languages without prior knowledge, while others may exploit it to craft deepfake videos that tarnish an individual’s reputation.
Navigating the Risks of AI: From Identity Theft to Deepfakes
Chereshev holds the belief that we currently need AI to defend us against other AI threats. With the rapid development of neural networks, new cybersecurity dangers are emerging. Personal or corporate digital footprints—such as search histories or social media posts—are potentially vulnerable to hackers’ whims.
For instance, people’s inquiries to conventional AI systems like ChatGPT become accessible to its creators, as openly described in the license agreements. Such AI can assign identifiers to individuals and collect comprehensive data about their activities.
AI Cyberthreats: Preparing for AI-Induced Scams and Frauds
Evgeny explained that a so-called “unlocked” version of generative AI lacks safeguard restrictions. With such tools, individuals with no technical expertise in hacking could attempt to commit cybercrimes, like bank fraud or stealing credit card information.
Phishing evolved into a sophisticated threat with deep phishing. In an illustrative scenario, a scammer, equipped with considerable knowledge about the victim, engages in a convincing phone call that could lead to malicious financial transactions.
A recent example of AI’s deceptive capabilities was a convincing but fake photograph of singer Katy Perry at the Met Gala 2024, which spread across the internet before being identified as a neural network’s creation.
Scammers have also used AI to compile compromising fake videos to extort victims who fear they cannot prove the footage was fabricated.
Defending Against Emerging Cyber Threats
Combating these threats is challenging, but Chereshev and the MTS Innovation Center team believe strides can be made. The company has committed up to 15% of its profits to R&D technologies, emphasizing AI. Their recent development, Membrana, is a service designed to shield device data from trackers while browsing the web.
Moreover, the AI technology behind virtual doppelgängers isn’t just used for deception. Projects like Altego by MTS have utilized it to create digital avatars capable of representing individuals in video conferences or instantly translating languages.
Digital Assistants: The Future of Telecommunications
Chereshev noted a growing societal acceptance of digital assistants. Initial skepticism gave way to routine use, with analog tools like Yandex’s Alice demonstrating how these technologies will inevitably become integral to people’s lives, handling mundane tasks from administrative work to shopping.
However, this integration isn’t without downsides, such as “digital amnesia.” As reliance on digital assistants grows, cognitive skills, such as memory retention and critical thinking, could diminish. Chereshev himself practices “digital detox” to counteract gadget dependency and advises others to consider it.
AI: Our Innovative Ally or a Cyber Adversary?
AI’s dual nature is most evident in its applications: it can serve as an invaluable ally in innovation or become an adversary by facilitating cyber threats. Breakthroughs in machine learning, natural language processing, and computer vision are enabling unprecedented progress across sectors—medicine, finance, transportation, etc. AI solutions can streamline complex tasks, enhance productivity, and even create life-saving technologies.
The Cybersecurity Arms Race: AI vs. AI
The rapid evolution of AI is a double-edged sword, particularly within cybersecurity landscapes. AI systems can model normal user behavior and detect anomalies indicating a cyber threat, or they can be utilized to launch sophisticated phishing campaigns that mimic genuine communications with alarming accuracy. The sophistication of these threats requires a parallel evolution in AI-driven security measures.
AI’s Ethical Quandaries
The ethics of AI are a simmering pot of debate, with deepfakes raising issues about consent, authenticity, and trust in digital media. The ability of AI to generate convincing false content poses threats to personal reputations, and as AI becomes more refined, distinguishing real from fake becomes even harder.
Aiding and Abetting Cybercrime
The potential for AI to empower individuals with malicious intent extends beyond deepfakes to encompass a range of cybercrimes. From automated hacking tools to bots capable of carrying out scams on a massive scale, the misuse of AI could result in significant financial and data losses.
Emerging Defense Strategies in AI
AI’s defensive capabilities are also advancing, with new software like MTS’s Membrana aiming to shield users from data tracking and provisions for detecting deepfakes and other AI-generated falsehoods. Investment in AI to combat AI threats is a sign of the proactive measures being undertaken by the tech industry.
The Price of Convenience: Digital Amnesia and Dependency
The debate over digital assistants concerns not just privacy and data security but also psychological and cognitive impacts. As AI handles increasing aspects of daily life, there is a risk that human cognitive abilities could atrophy, leading to ‘digital amnesia,’ where reliance on technology significantly lessens memory capabilities.
Conclusion
Overall, AI technology promises great advances while presenting new risks that society must carefully navigate. As AI takes on more roles in our personal and professional lives, balancing the scales between embracing innovation and mitigating potential hazards remains a crucial challenge.
For more insights about AI innovations and cyberthreats, consider visiting trusted resources such as MIT Technology Review for their articles on technology, or IEEE Spectrum for more on the latest in engineering and applied sciences.