GenAI Security Researcher

Location: Ramat Gan, Israel

Company

hirify.global, operating through Alice, provides trust, safety, and security solutions across the AI lifecycle, including model hardening evaluations, pre-deployment red-teaming, runtime guardrails, and drift detection.

What you will do

• Conduct hands-on adversarial research on frontier text, code, multimodal, and reasoning models to identify jailbreaks, prompt-injection vectors, guardrail bypasses, and policy-evasion techniques.
• Research agentic attack surfaces involving tool use, MCP and plugins, browsing and computer-use agents, coding agents, and multi-agent pipelines.
• Assess offensive cyber capabilities such as vulnerability discovery, exploit development, malware, lateral movement, privilege escalation, credential abuse, SSRF, injection, and supply-chain attacks in LLM-based systems.
• Build adversarial datasets, attack taxonomies, evaluation scenarios, and automated attack-generation and testing tooling using Python, LLM-assisted tools, and agent frameworks.
• Track model releases, academic research, published attacks, and real-world abuse, turning them into actionable research directions.
• Communicate findings through reports, write-ups, talks, responsible disclosure, and collaboration with red-teamers, data scientists, engineers, customer teams, and research partners.

Requirements

• At least 5 years of hands-on cybersecurity research experience in offensive security, red teaming, vulnerability research, reverse engineering, detection research, or exploit development.
• At least 1 year of hands-on experience attacking or evaluating GenAI systems, including jailbreaks, direct or indirect prompt injection, guardrail bypass, or LLM red teaming.
• Solid understanding of modern LLM operation and deployment, including training and alignment basics, RLHF, safety tuning, system prompts, context windows, tool calling, RAG, and agent architectures.
• Strong attacker mindset with the ability to model unfamiliar systems, identify trust boundaries, and find creative ways to break them.
• Strong coding skills and comfort building custom tooling and automation.
• Ability to work independently on open-ended research problems from hypothesis through proof of concept and clear write-up.
• Excellent written and verbal communication in English.

Nice to have

• Published AI security research, CVEs, conference talks, or public jailbreak and prompt-injection write-ups.
• Experience red teaming production AI agents or AI-powered applications, or participating in AI bug bounty programs.
• Background in threat intelligence, incident response, or detection engineering.
• Familiarity with OWASP Top 10 for LLMs, MITRE ATLAS, and NIST AI RMF.
• Experience building adversarial benchmarks or LLM evaluation harnesses.
• Working knowledge of ML fundamentals or open-weight models, including fine-tuning, inference, or interpretability.
• Offensive security certifications such as OSCP, OSWE, OSED, OSCE3, or GREM, or a strong CTF track record.

Culture & Benefits

• Full-time role based in Ramat Gan, Israel.
• Hands-on research with frontier models and agents, with opportunities to publish internally and externally.
• Collaboration with red-teamers, data scientists, engineers, research partners, and customer teams.
• Work focused on improving the security of AI systems used by major AI and technology platforms.