Job Title: Threat Research Analyst – AI Security

Department: Security Research

Experience: 3–6 years (flexible based on relevant skills)

Job Type: Full-time

Role Overview

We are looking for a Threat Research Analyst specializing in AI security to investigate emerging threats targeting AI applications, large language models (LLMs), AI agents, and the broader AI ecosystem.

The role focuses on identifying and analyzing adversary behaviors, researching AI-specific attack techniques, tracking emerging vulnerabilities and threat campaigns, and translating research findings into actionable threat intelligence. The analyst will work closely with the Security Content Research team to convert research into detection use cases, threat hunting opportunities, and security recommendations.

The ideal candidate should have a strong cybersecurity research foundation, demonstrated analytical skills, and a keen interest in AI security, adversarial techniques, and the evolving risks associated with enterprise AI adoption.

Key Responsibilities

1. AI Threat Intelligence and Adversary Research

- Research emerging threats targeting AI applications, LLM platforms, AI agents, AI development environments, and supporting infrastructure.
- Monitor threat actor activity, security advisories, vulnerability disclosures, research publications, and incident reports relevant to AI security.
- Investigate attacks involving AI application abuse, prompt injection, model manipulation, sensitive data exposure, malicious tool execution, and compromised AI workflows.
- Track adversary tactics, techniques, procedures, and indicators of compromise relevant to AI-enabled environments.
- Analyze how established attack techniques are adapted to exploit AI applications, agent frameworks, and AI-integrated enterprise systems.
- Maintain structured research notes, threat profiles, attack timelines, and technical evidence.

2. AI and Agentic Application Security Research

- Investigate security risks associated with LLMs, retrieval-augmented generation (RAG), AI plugins, external tools, and autonomous or semi-autonomous AI agents.
- Research risks involving excessive agent permissions, insecure tool invocation, indirect prompt injection, identity misuse, data leakage, and unauthorized actions.
- Analyze AI service and agent telemetry to identify suspicious behaviors and potential attack patterns.
- Research security risks across AI platforms and services, including ChatGPT, Anthropic Claude, Google Gemini, Microsoft Copilot, and relevant cloud AI platforms.
- Examine threats to AI development and execution environments, including coding assistants, command-line AI tools, model integrations, and agent frameworks.
- Identify gaps in available telemetry and recommend additional data sources needed for AI threat detection and investigation.

3. Vulnerability and Exploit Research

- Monitor vulnerabilities affecting AI applications, AI infrastructure, agent frameworks, model-serving components, and related software.
- Analyze vulnerability disclosures, proof-of-concept exploits, technical advisories, and published attack research.
- Assess exploitability, attack prerequisites, potential impact,



affected technologies, and defensive implications.
- Reproduce or safely validate relevant attack scenarios in controlled environments where feasible.
- Identify observable behaviors, artifacts, and indicators that can support threat hunting and detection development.
- Coordinate with the broader research team to prioritize vulnerabilities and emerging threats based on risk and operational relevance.

4. Threat Hunting and Detection Collaboration

- Convert threat research findings into actionable hunting hypotheses, investigation procedures, and detection requirements.
- Collaborate with Content Research to develop behavioral detections and AI security use cases based on validated research.
- Map relevant adversary behaviors to MITRE ATT&CK;, MITRE ATLAS, and applicable OWASP AI security guidance.
- Identify useful indicators, event fields, behavioral patterns, and log sources for investigating AI-related threats.
- Support the development of threat intelligence reports, IOC publications, security advisories, and technical research.
- Provide technical context to help distinguish malicious behavior from legitimate AI application and agent activity.

5. Threat Intelligence Reporting and Dissemination

- Produce technically accurate threat intelligence reports, vulnerability analyses, research briefs, and security advisories.
- Explain attack chains, affected technologies, observed behaviors, potential business impact, and actionable mitigation measures.
- Develop research findings suitable for security practitioners, SOC analysts, threat hunters, and security leadership.
- Support internal briefings, technical presentations, and external research publicationactivities.
- Contribute to the organization's AI security research roadmap by identifying emerging topics and high-value research opportunities.

6. Research Frameworks and Knowledge Development

- Track developments in MITRE ATLAS, OWASP Top 10 for LLM Applications, OWASP Agentic Applications, and other relevant AI security frameworks.
- Maintain structured mappings between AI attack techniques, threat scenarios, telemetry requirements, and defensive opportunities.
- Build reusable research artifacts, technical references, attack scenario documentation, and threat hunting guidance.
- Evaluate emerging AI security tools, research methodologies, and open-source projects relevant to threat investigation.
- Identify opportunities to automate threat monitoring, research collection, data enrichment, and analysis using Python or other scripting tools.

Required Skills and Qualifications

- Bachelor's degree in Computer Science, Information Security, Information Technology, or a related discipline, or equivalent practical experience.
- 3+ years of relevant experience in threat intelligence, security research,



incident response, vulnerability research, malware analysis, or threat hunting.
- Robust understanding of cybersecurity fundamentals, including network security, endpoint activity, authentication, identity, privilege abuse, and common attack techniques.
- Demonstrated ability to investigate technical security issues using logs, public research, security advisories, and other technical evidence.
- Understanding of threat actor TTPs, indicators of compromise, attack chains, and threat intelligence analysis.
- Familiarity with LLMs, AI applications, AI agents, or the security risks of AI-integrated systems; demonstrated practical interest or project experience is essential.
- Ability to evaluate security research critically, validate technical claims, and distinguish confirmed findings from assumptions.
- Strong technical writing and research communication skills.
- Ability to translate complex technical findings into clear, actionable defensive recommendations.

Preferred Qualifications

- Hands-on experience researching prompt injection, indirect prompt injection,

jailbreaks, insecure AI tool use, data exfiltration, or agent security.
- Familiarity with MITRE ATLAS, OWASP Top 10 for LLM Applications, OWASP Agentic Applications, or related AI security frameworks.
- Experience analyzing AI service logs, application audit events, API activity, identity events, or agent execution telemetry.
- Familiarity with Python, scripting, API analysis, and automated research workflows.
- Experience with vulnerability analysis, proof-of-concept validation, malware analysis, or controlled attack simulation.
- Understanding of cloud security, SaaS platforms, identity and access management, and enterprise AI deployments.
- Experience writing threat intelligence reports, vulnerability advisories, technical blogs, or security research publications.
- Familiarity with threat intelligence platforms, IOC enrichment, and structured threat intelligence formats.

Expected Outcomes

- Continuous research coverage of emerging threats targeting AI applications, LLMs, and agentic systems.
- Timely identification and analysis of relevant vulnerabilities, attack techniques, and threat campaigns.
- Production of actionable AI threat intelligence reports, research briefs, and security advisories.
- Development of validated AI threat hunting hypotheses and detection requirements.
- Improved coverage of AI-specific attack techniques through collaboration with the Content Research team.
- Identification of telemetry gaps and recommendations for new AI security integrations and data sources.
- Maintenance of structured AI threat research mapped to relevant industry frameworks.

Ideal Candidate Profile A research-oriented cybersecurity practitioner who can independently investigate emerging threats, validate technical findings, understand adversary behavior, and communicate actionable defensive insights. The candidate should be able to connect AI-specific attack techniques with real-world telemetry, threat hunting, and detection opportunities while distinguishing evidence-backed findings from theoretical risks.

📌 Threat Research Analyst – AI Security (Pune)
🏢 Gurucul
📍 Pune