Articles by tag: prompt-injection

Attacks on LLM in Business: Vulnerabilities of Local Models

How Attackers Extract CRM Data Through LLM? Analysis of Real Attacks on Local and Cloud Models. Protection of SMB Solutions.

LLM Security: Evolution from Firewall to GAF | In-Depth Analysis

How modern threats to LLM require multi-level protection. We analyze incidents of 2025-2026 and the transition to Generative Application Firewall. Learn how to protect AI systems.

LLM Vulnerability: Bypass via assistant prefill

Learn how sockpuppeting bypasses protections of 11 language models with one line. Risk analysis, protection measures, and business implications. Read attack details and recommendations.

Marusya and Salut Vulnerabilities: Bypassing Filters

Analysis of Technical Loopholes in Voice Assistants Marusya and Salut. Prompt injection via Choices, Memory, and Reminders. Protection Recommendations for LLM Developers.

Protocol for LLM: fighting hallucinations

Learn how the 1100-line protocol eliminates hallucinations in Qwen and DeepSeek. Tests show 100% protection from injections and errors. Apply for code and facts — launch instructions.

Protection of agent applications OWASP Top 10

Analysis of OWASP Agentic Top 10 2026 and Lethal Trifecta for protecting AI agents. Countermeasures against goal hijack, RCE, memory poisoning. Practical roadmap for middle/senior dev.