Practical content from industry professionals on digital forensics · data recovery · cyber security · KVKK compliance.
A public chatbot is a door both the customer and the attacker talk to. A table of common chatbot risks, data leakage and brand risk, instruction input separation and access minimization defense and evidence based testing with KAOS.
Read moreAI learns from past code and the code it produces can carry insecure patterns. A table of common code risks, the developer's over trust problem, security testing, the supply chain dimension (nonexistent package) and evidence based auditing with KAOS.
Read moreWhen an AI system is attacked the symptom is often not a crash but silently wrong decisions. A table of AI specific incident types, response steps (contain, scope, root cause, safe version) and evidence preserving examination with KAOS.
Read morePrivacy preserving ML trains a model without disclosing sensitive data. The risk of the central data pool, a table of methods (federated learning, differential privacy), the privacy accuracy balance, KVKK compliance and model privacy assessment with KAOS.
Read moreIf an AI agent is given more tools and autonomy than it needs, a manipulation can use these privileges for the attacker. A table of the three dimensions of excessive agency, from manipulation to destructive action, least privilege and human approval defense and the KAOS approach.
Read moreAI can convincingly imitate voice and face; 'I recognized their voice' is no longer proof of identity. Why voice and image are not enough, executive fraud, a layered identity verification table (second channel, code word) and assessment with KAOS.
Read moreAI catches fake transactions and account takeovers in real time by learning deviation from normal behavior. The difference from rule based, a table of patterns it catches, the false positive and miss balance, human oversight and the KAOS approach.
Read moreWhen an AI says 'this is an attack', the security team must see the reason. The black box problem, a table of why explainability is critical, why it is a must in forensics, the evidence based approach and the explainable design of KAOS.
Read moreA model you download can carry a hidden backdoor; the AI supply chain is a new risk. A table of supply chain risks, verifying source with provenance, testing the downloaded, component inventory (SBOM) and assessment with KAOS.
Read moreModel security is protecting a machine learning model itself against stealing and extracting the data inside it. A table of attacks against the model, model inversion and privacy, access and query limit defense and assessment with KAOS.
Read moreA cyber security AI is strong at scaled scanning and noise filtering, needs the human for business context and real exploitation. Capability and limit tables, a framework to evaluate a tool (false positives, local processing, honesty) and the KAOS approach.
Read moreAI transforms the SOC by filtering the alert pile and reducing noise. The SOC's real problem is alert fatigue, a table of what AI changes, what it does not do, setup models (MDR) and human supervised monitoring with KAOS.
Read more