_edited.png){kind=small}
BUILD YOUR PATH
OPPORTUNITIES ARE HERE AND THEY ARE MEANT TO BE TAKEN
WOULD YOU?
AI Red Teamer
Full Time
Information Technology AI
México
Salary Range: US$55,000.00 to US$85,000.00* per Annually
*Depending on experience, capabilities & performance
CA-IT-IA-1290
Job Description
Key Responsibilities:
Conduct penetration testing on Claude-based apps and identify logic flaws in AI reasoning.
Adversarial Campaign Design: Create and execute structured attack plans targeting specific risk categories (e.g., PII leakage, harmful content generation, or tool-use hijacking).
Jailbreak Discovery: Proactively find and document new methods to bypass Claude’s safety protocols (e.g., via role-play, "Crescendo" multi-turn attacks, or complex encoding).
Vulnerability Reporting: Translate successful attacks into clear, technical remediation tickets for the LLM Engineers.
Automated Red Teaming: Build and maintain pipelines that use other LLMs to automatically generate millions of adversarial inputs against our production models.
Benchmarking Safety: Measure the "Attack Success Rate" (ASR) across different model versions (Claude 3.5 Sonnet vs. Haiku) to guide deployment decisions.
Key Activities:
Jailbreaking attempts, data extraction simulations, and reporting vulnerabilities.
Prompt Injection Attacks: Testing both Direct (user-to-model) and Indirect (via RAG or web-search) prompt injections to gain unauthorized control.
Data Extraction Simulations: Attempting to trick the model into revealing system prompts, internal knowledge base fragments, or personal user data.
Multi-turn Escalation: Conducting "Crescendo" attacks where the model’s helpfulness is exploited over several turns to gradually wear down its refusal mechanisms.
Tool-Use Exploitation: Attacking the "function calling" capabilities to see if Claude can be manipulated into executing malicious code or unauthorized API calls.
Adversarial Training Data Generation: Curating datasets of "failed" prompts to help harden the model via fine-tuning or better system instructions.
Academic Skills and Qualifications
Degree in Cybersecurity or Computer Science: A Bachelor’s or Master’s degree with a heavy focus on Information Security, Pen-Testing, or Adversarial Machine Learning.
Offensive Security Methodologies: Academic mastery of the MITRE ATLAS™ (Adversarial Threat Landscape for Artificial-Intelligence Systems) framework.
LLM Vulnerability Research: Specialized knowledge in "Jailbreaking" theory, Prompt Injection (Direct and Indirect), and Virtual Prompt Attacks.
Advanced Natural Language Processing (NLP): Understanding how Transformers (like Claude) tokenize and process text to find "semantic bypasses" in safety filters.
Adversarial Machine Learning: Theoretical knowledge of Data Poisoning, Model Evasion, and Membership Inference attacks against RAG (Retrieval-Augmented Generation) systems.
Ethical Hacking & Penetration Testing: Academic grounding in traditional network and application security (OWASP Top 10 for LLMs) to ensure the infrastructure around the AI is as secure as the model itself.
Scripting for Automated Exploitation: Advanced proficiency in Python and Bash to build custom "fuzzing" tools that send thousands of adversarial queries to the AI.
Social Engineering & Cognitive Psychology: Understanding of human-centric attacks, specifically how AI can be manipulated through "Persona Adoption" or social pressure tactics.
Privacy & Extraction Math: Knowledge of "Model Inversion" math to test if the AI can be forced to reveal its private training data or sensitive Real Estate listings it shouldn't have access to.
Game Theory: Ability to use game-theoretic models to anticipate attacker behavior and design "defense-in-depth" strategies.
Forensic Analysis of AI Outputs: Academic training in identifying "fingerprints" of model manipulation and trace analysis in LLM logs.
Regulatory & Ethical Standards: Deep understanding of the legal boundaries of "White Hat" hacking and the ethics of reporting vulnerabilities in a professional Real Estate environment.
Key Skills and Qualifications:
Ethical Hacking certification (CEH/OSCP), deep understanding of LLM tokens and biases.
Offensive Mindset: Proven ability to think "outside the box" to bypass logic-based constraints.
Technical Tools: Proficiency with adversarial ML libraries like Giskard, PyRIT (Python Risk Identification Tool), or Promptfoo.
Anthropic Knowledge: Deep understanding of Claude’s Constitutional AI and how it differs from RLHF-based models.
Cybersecurity Foundations: Knowledge of the OWASP Top 10 for LLMs and traditional MITRE ATT&CK frameworks.
Coding: Advanced Python for automating attack scripts and analyzing token-level vulnerabilities.
Certifications: OSCP (Offensive Security Certified Professional) or specialized AI Security certifications are highly valued.
Work Experience:
3+ years in Penetration Testing.
Daily Job Schedule:
The daily job schedule will follow the standard working hours of the company, typically from 9:00 AM to 5:00 PM. However, occasional flexibility may be required to meet project deadlines or collaborate with remote team members.
What we offer!
At our organization, we work hard to establish a welcoming environment that draws in and keeps great people. We are aware of the importance of a supportive workplace environment for our employees' success and happiness. Here is a summary of the ideal workplace culture we provide to attract all job hopefuls:
-
An inclusive and diverse culture is important to us because we think that different viewpoints and experiences foster better creativity and problem-solving. Every employee feels accepted, valued, and empowered to offer their special thoughts and experiences because we promote an inclusive atmosphere.
-
We encourage teamwork and collaboration because we know that when people work together, they typically get the best results. To create a friendly and supportive workplace, we promote open communication, idea exchange, and cross-functional cooperation.
-
Opportunities for Growth and Development: We place a high priority on our workers' continued development. To improve their skills and expertise, we provide a range of learning and development programs, mentorship opportunities, and ongoing training. We give employees a clear career path and aid them in reaching their professional objectives.
-
Work-Life Balance: We are aware of how crucial it is to keep a positive work-life balance. We encourage employees to take time off to recover and pursue personal interests and we provide flexible work schedules, remote work opportunities (where appropriate), and other benefits. We think that having a healthy work-life balance makes people happier and more effective.
-
Recognizing and rewarding hard work and contributions from employees is important to us. We offer benefits and attractive salary packages that are in line with industry standards, and we have a structured recognition program that recognizes excellent achievement.
-
Employee Support: We place a high priority on our employees' support and well-being. We provide access to services for both physical and mental health as well as comprehensive health insurance policies, wellness initiatives, and programs. We encourage staff to express their concerns or ask for help when necessary and maintain an open-door policy.
-
Environment for Innovation and Creativity: We promote an atmosphere for innovation and creativity. We give workers the freedom, autonomy, and resources they need to try out novel concepts, innovative ideas, and cutting-edge technologies. We support questioning the status quo and making continual improvements to our procedures and products.
-
Social responsibility and sustainability are important to us because we want to have a beneficial influence on both people and the environment. We offer volunteer opportunities, sustainability programs, and corporate social responsibility activities so that staff members can support worthwhile causes and change the world.
-
Transparent Communication: We support open lines of communication and educating staff on new initiatives, objectives, and tactics. We have regular town halls, team meetings, and offer avenues for comments and feedback. We value candid conversation and welcome employee ideas and input.
-
Fun and Engaging Activities: We are committed to fostering a positive work environment. We plan social events, team-building exercises, and joint milestone and achievement celebrations. We support an upbeat, welcoming environment where humor and creativity flourish.
Benefits
Health & Wellness
Fitness & Recreation
Health Insurance
Life Insurance
Continuing Education
Training & General Education
Foreign Language
Family
Flexible Scheduling
Child Care
Finance
Retirement Plan
Bonus / Compesation
Free Time
Paid Off Time
Sick Leave
Law Enforcement & Veteran Benefits
Special Discounts
Special Learning Programs
DISCLAIMER
We are proud to foster a workplace free from discrimination. We strongly believe that diversity of experience, perspectives, and background will lead to a better environment for our employees and a better product for our users and our creators. This is something we value deeply and we encourage everyone to come be a part of changing the way the world.
IMPORTANT: earnings and legal disclaimers Carlos Ayala is an Internet marketing and security professional and his results are not typical. Your experiences are not a guarantee that you will earn money. You can do more, less or the same or nothing at all. This is purely educational. No income is guaranteed.
Only serious and ambitious entrepreneurs please apply.
Prices are subject to change without notice. No refunds will be allowed for tickets at any price level. If for some reason you cannot attend the events that are organized, you will receive full credit for your investment in my store (iamcarlosayala.com). There is no risk for your investment with me today.