Resume Example & Template

AI Red Teamer Resume Example

For security people breaking language models — a field with real demand and almost no CV conventions.

What is a AI Red Teamer?

An AI Red Teamer attacks AI systems on purpose. That covers prompt injection and jailbreaks, but the serious work is systemic: indirect injection through retrieved content, abuse of an agent’s tools, data exfiltration, and training-data extraction. The output is a report engineering teams can act on, and most real findings require a design change rather than a filter.

It is one of the most natural pivots in the industry. Offensive security already teaches the mindset — assume the input is hostile, chain small weaknesses into a real one, prove impact rather than theory. What changes is the attack surface: models follow instructions from anywhere in their context, which breaks the boundary between data and command.

Demand is strong wherever AI touches money, health or personal data, and it will grow as agents gain the ability to take actions rather than only produce text. The talent pool is thin because it needs both security experience and genuine familiarity with how models behave.

Key skills for a AI Red Teamer resume

prompt injection
indirect prompt injection
jailbreaks
agent security
tool-use abuse
data exfiltration
model extraction
adversarial testing
threat modelling
penetration testing
OWASP LLM Top 10
MITRE ATLAS
NIST AI RMF
responsible disclosure
security reporting

AI Red Teamer resume example

Sofia Almeida

AI Red Teamer & Security Engineer

Lisbon, Portugal

Summary

Security engineer turned AI red teamer. 8 years in offensive security, 3 breaking language models — prompt injection, jailbreaks, tool-use abuse and agent exfiltration. Found the indirect injection chain that changed how a client handles retrieved content, and built the adversarial suite run on every release.

Experience

Lead AI Red Teamer · Arclight Security

Feb 2024 – Present

Lead adversarial testing of LLM products for financial and healthcare clients, covering injection, jailbreaks, tool abuse and data extraction.
Discovered an indirect prompt-injection chain enabling exfiltration via an agent’s tool calls; drove the client’s redesign of untrusted-content handling.
Built an automated adversarial suite of 300+ attack patterns, run on every model release.
Write client-facing reports and brief engineering teams on fixes — most findings need a design change, not a filter.

Senior Penetration Tester · Arclight Security

May 2020 – Jan 2024

Led web, API and cloud penetration tests for regulated clients and ran internal training on emerging attack surfaces.
Built the tooling that became the basis for the AI red-team practice.

Education

MSc Information Security

Instituto Superior Técnico · 2015 – 2017

BSc Computer Engineering

University of Porto · 2012 – 2015

Certifications

OSCP — Offensive Security Certified Professional
CISSP

Skills

AI Security: Prompt Injection · Jailbreaks · Agent Security · Tool-Use Abuse · Data Exfiltration

Offensive Security: Penetration Testing · Threat Modelling · OWASP LLM Top 10 · API Security

Frameworks: MITRE ATLAS · NIST AI RMF · Responsible Disclosure

Use this as your starting point

How to write a AI Red Teamer resume that stands out

Lead with a finding that changed a design. "Found an indirect injection chain that led to a redesign of untrusted-content handling" says more than any list of techniques.
Separate the systemic from the trivial. Anyone can produce a jailbreak; showing you found an exfiltration path through an agent’s tools demonstrates you understand the architecture.
Name the frameworks — OWASP LLM Top 10, MITRE ATLAS, NIST AI RMF. They are screened for literally and they show you can work inside an existing security programme.
Keep your traditional security credentials visible. OSCP, CISSP or equivalent still carry weight and reassure a hiring manager you understand disclosure and reporting discipline.
Show the reporting side. Red teaming that engineers cannot act on has no value, and communication is what separates senior from junior here.
If you are pivoting in, evidence counts more than titles: public write-ups, CTF work on AI challenges, or a documented responsible disclosure will move you past applicants with neither.

AI Red Teamer resume — FAQ

Can I move into AI red teaming from traditional penetration testing?

Yes, and it is the most common route. The mindset transfers directly. What you need to add is hands-on familiarity with how models handle context, tools and retrieval — which you can build on your own before applying.

Do I need to understand machine learning deeply?

You need to understand behaviour more than mathematics. Knowing why a model follows an instruction buried in a retrieved document matters far more than being able to derive backpropagation.

What can I show publicly without breaching disclosure rules?

Techniques, methodology and generic findings, never client specifics. Many practitioners publish on public models or their own deployments, which is entirely acceptable evidence and often how people get hired.

Is this the same as AI safety work?

They overlap but differ in aim. Safety research studies harmful model behaviour broadly; red teaming targets a specific deployed system to find exploitable weaknesses. Some roles combine both.

Ready to land your dream job?

Join all the job seekers who have successfully built their resumes and advanced their careers with CopilotResume.

Build Your Resume Now Browse Templates

Transparent and cost-effective pricing plans.