Sandra S.

Ethical Hacker | Ex-CISO | Security Expert

My professional background is defined by the combination of offensive security operational experience and strategic-level information security governance.

I built my expertise on the technological frontlines through vulnerability assessments of critical systems, and later integrated this operational knowledge into enterprise security strategies and the defense protocols of complex infrastructures. My core domain is the resilience of IT ecosystems, with a particular focus on the security validation of artificial intelligence architectures and the analysis of emerging algorithm-based threat vectors.

Academic and Professional Pillars

Early professional orientation toward LLM security(2018):

Even prior to the widespread adoption of generative models, I was extensively engaged with the cybersecurity aspects of conversational agents.

I was particularly interested in the risks arising from the emotional dynamics of AI-driven interactions, as well as the integrity of early language models and the vulnerabilities of human–machine interfaces. I formally proposed this topic as the focus of my thesis research in 2018. During the preparatory phase, I conducted regular exploratory interactions with contemporary conversational systems, the insights from which further reinforced the urgency and significance of this field.

Research on the future of cybersecurity:

During my master’s studies, well before the widespread adoption of LLM technologies, I chose “AI-based cyberattacks” as my research domain. In my thesis, I examined the role of AI systems in cyber warfare, with a focus on automated attack mechanisms and the risks of predictive models.

My research examined, among other topics, the potential for embedding malicious code within neural networks, as well as AI-based, context-dependent, and behavior-driven decision-making mechanisms that enable precise and targeted malware activation.

Enterprise AI security: From February 2023, at the beginning of the global expansion of generative AI, I had the opportunity to become a member of the first dedicated AI task force within a highly regulated enterprise environment. In this sector, establishing appropriate security frameworks was critical, as innovative AI solutions had to simultaneously meet the strictest cybersecurity, data protection, and regulatory requirements.

As a cybersecurity expert, I was responsible for the secure and compliant integration of AI technologies into the organization’s critical business processes.

My responsibilities covered the entire AI lifecycle from a security perspective: from internally developed solutions, through comprehensive risk analysis of third-party models, to ensuring the integrity of training data and conducting cybersecurity validation of models.

In my current advisory work, I focus on enabling partners to introduce AI solutions that enhance operational efficiency by establishing the right balance between innovation and security. This is achieved within optimized frameworks that avoid unnecessary costs while aligning security levels with real-world risks.

Mission of the Publication

I created this platform because I see that the AI revolution brings both unprecedented opportunities and risks never seen before.

My goal is to build a comprehensive, structured knowledge space capable of capturing, organizing, and making interpretable the full spectrum of AI security.

I aim to create a conceptual and informational map that not only summarizes the rapidly expanding body of knowledge but also organizes it into a coherent framework, enabling deeper understanding and critical analysis of the field.

I believe that a broad and in-depth understanding of how AI systems operate, their limitations, and their risk profiles forms the foundation of our collective responsibility. The more stakeholders — researchers, decision-makers, and society at large — are able to interpret the direct and indirect impacts of artificial intelligence, the greater the chance that technological advancement will be not only efficient, but also consciously guided from ethical and security perspectives.