Artificial intelligence is rapidly becoming part of everyday life. It helps us work, communicate, create content and even defend against cyber threats. But as AI evolves from a helpful assistant into a technology capable of acting autonomously, influencing decisions and reshaping online experiences, new questions emerge.
At Kaspersky HORIZONS 2026 in Rome, cybersecurity researchers, AI experts, policymakers and educators gathered to explore those questions. One message echoed throughout the event: the future of cybersecurity is no longer only about protecting systems and data. It is increasingly about protecting trust, identity and human autonomy.
1. AI agents are creating a new attack surface
Cybercriminals have always exploited trust. Today, they are exploiting trust in AI.
Research presented at HORIZONS revealed that between January and early May 2026, Kaspersky technologies detected more than 92,000 malware attacks disguised as popular AI services. Fake versions of ChatGPT, Claude, Gemini and other emerging AI tools were used to trick users into downloading malicious software.
The threat extends beyond fake applications. Attackers are increasingly targeting AI supply chains, open-source libraries and interconnected AI ecosystems, where a single compromised component can affect thousands of users or organizations.
As Dmitry Galov, Head of Kaspersky’s Global Research and Analysis Team (GReAT) Russia and CIS Unit, explained:
“The introduction of AI agents into enterprise environments changes the nature of trust itself. Every automated action becomes part of a wider chain of systems and data exchanges, which means security is no longer just about protecting endpoints – it is about controlling how intelligence, permissions, and decisions propagate across interconnected AI-driven processes.”
Dmitry Galov, Head of Kaspersky’s Global Research and Analysis Team (GReAT) Russia and CIS Unit, at Kaspersky HORIZONS 2026
As organizations embrace AI, securing these intelligent ecosystems is becoming just as important as protecting traditional networks.
2. Automation needs human oversight
AI promises greater efficiency, faster threat detection and improved automation. Yet autonomy also introduces new risks.
Experts discussed how AI systems can be manipulated through prompt injection, poisoned datasets, malicious plugins and compromised supply chains. When automated systems operate at scale, mistakes and vulnerabilities can spread quickly.
The challenge is no longer only technical resilience, but maintaining accountability.
As Luana Lo Piccolo, Senior Advisor on Tech Law, AI Governance and Digital Global Affairs, noted:
“As AI systems evolve from assistants into autonomous actors, the challenge is no longer only technical resilience, but accountable autonomy.”
Luana Lo Piccolo, Senior Advisor on Tech Law, AI Governance and Digital Global Affairs, at Kaspersky HORIZONS 2026
The more organizations rely on AI-driven decisions, the more important governance, transparency and human supervision become.
3. AI is beginning to understand human behavior
One of the most fascinating discussions at HORIZONS focused on cognitive AI — systems designed to interpret human behavior, anticipate decisions and adapt interactions accordingly.
While current technologies cannot read minds, they already influence choices through recommendation engines, personalization algorithms and behavioral prediction.
Researchers identified four emerging risks:
- More sophisticated social engineering attacks
- Large-scale manipulation of public opinion
- Predictive profiling and targeted abuse
- Future convergence between brain-computer interfaces and connected devices
Noushin Shabab, Lead Security Researcher at Kaspersky GReAT, believes these risks deserve attention today, not tomorrow:
“Although cognitive AI is still at an early stage and far from mass adoption, it is developing rapidly. Advanced human-AI interaction models are still expected to become significantly more widespread in the coming decades. As adoption will grow, so too will the associated risks – and when it happens, we need to be prepared.”
Noushin Shabab, Lead Security Researcher at Kaspersky GReAT, at Kaspersky HORIZONS 2026
The conversation highlighted a growing reality: cybersecurity may increasingly involve protecting not only devices, but also human decision-making.
4. Human autonomy is becoming a cybersecurity issue
As AI systems become better at understanding behavior, concerns are growing about their ability to influence it.
Social platforms already use algorithms to shape what users see, read and engage with. Future AI systems could become even more effective at tailoring information, predicting reactions and nudging decisions.
For journalist and responsible AI educator Teresa Potenza, this is where the debate becomes bigger than technology itself:
“The real risk of cognitive AI is that it shapes our minds, quietly and at scale. We have learned that systems optimized for engagement erode judgment.”
She added:
“Technology must serve people, not the other way around. Autonomy is not just a privacy issue. It is a democracy issue.”
Teresa Potenza, journalist and responsible AI educator, at Kaspersky HORIZONS 2026
Protecting users in the AI era will require more than security controls. It will also depend on transparency, education and responsible governance.
5. Seeing is no longer believing
A live demonstration at HORIZONS challenged one of the most basic assumptions about identity online.
Researchers showed how facial recognition systems could still identify individuals even after generative AI dramatically altered their appearance through aging and rejuvenation effects. To human observers, the images often appeared to show completely different people.
Yet facial recognition algorithms successfully matched the modified images to their original identities.
According to Maher Yamout, Lead Security Researcher at Kaspersky GReAT:
“AI-generated facial transformations may preserve biometric identity even when human perception interprets the images as entirely different individuals. This creates new challenges for digital trust, identity verification, and fraud prevention in an era of rapidly evolving synthetic media.”
Maher Yamout, Lead Security Researcher at Kaspersky GReAT, at Kaspersky HORIZONS 2026
The findings highlight both the resilience of modern biometric systems and the growing challenges posed by deepfakes and synthetic identities.
The future of cybersecurity is ultimately human
Kaspersky HORIZONS 2026 showed that AI is reshaping cybersecurity in ways that extend far beyond malware and network protection.
As intelligent systems become more autonomous, persuasive and integrated into daily life, the focus is shifting toward protecting trust, identity and human autonomy.
The next cybersecurity frontier may not be our devices or networks — but our ability to trust what we see, what we know, and even how we make decisions in an increasingly AI-driven world.
