TeracCybersecurity
Offensive Cybersecurity Engineers: Build RL Environments for AI Evaluation
$150/hr
About this role, in Terac's words
We are running a long-term paid project to build and harden offensive cybersecurity reinforcement learning environments. We need experienced cybersecurity professionals to develop benchmarks that evaluate the capabilities of frontier AI models.
What you'll do
What you would do
- Design and build offensive cybersecurity reinforcement learning environments
- Harden existing benchmarks to withstand advanced AI interactions
- Develop rigorous evaluation criteria for frontier AI models
- Collaborate on a long-term hourly basis to iterate on security scenarios
Candidate background
What this role asks for
- Offensive Cybersecurity
- Reinforcement Learning
- AI Evaluation
- Penetration Testing
Read from the posting's own words: show the exact sentences
- Offensive Cybersecurity: “listed by the platform under required skills: Offensive Cybersecurity”
- Reinforcement Learning: “listed by the platform under required skills: Reinforcement Learning”
- AI Evaluation: “listed by the platform under required skills: AI Evaluation”
- Penetration Testing: “listed by the platform under required skills: Penetration Testing”
Quoted from selected sections of Terac's posting. The full posting may state other conditions. Read the full posting
A closer look
More context, when you need it.
Full job descriptionThe original posting, in Terac's words
Posted by Terac, reproduced here so you can judge the role before clicking. Original posting ↗
We are running a long-term paid project to build and harden offensive cybersecurity reinforcement learning environments. We need experienced cybersecurity professionals to develop benchmarks that evaluate the capabilities of frontier AI models.
What We're Researching
We're hiring offensive cybersecurity experts for a long-term paid project focused on AI evaluation. Our goal is to build and harden reinforcement learning environments that test the limits of frontier AI models. This work directly contributes to understanding and mitigating advanced security risks in emerging technologies. How It Works
You will work remotely on an hourly basis to design and implement complex cybersecurity benchmarks. Your tasks will involve creating robust reinforcement learning environments tailored for offensive security scenarios. You will continuously harden these environments to ensure they provide rigorous evaluations for AI systems. The project requires ongoing collaboration and iterative development to keep pace with new vulnerabilities. Who This Is For
We welcome offensive cybersecurity engineers, penetration testers, and security researchers with strong technical backgrounds. Ideal candidates have hands-on experience building secure environments or developing security benchmarks. Familiarity with reinforcement learning or AI evaluation frameworks is highly valued but not strictly required.
What you would do
• Design and build offensive cybersecurity reinforcement learning environments
• Harden existing benchmarks to withstand advanced AI interactions
• Develop rigorous evaluation criteria for frontier AI models
• Collaborate on a long-term hourly basis to iterate on security scenarios
Who this is for
• Professional experience in offensive cybersecurity or penetration testing
• Strong background in building or hardening technical environments
• Comfortable working on long-term hourly remote projects
• Interest in AI safety and model evaluation
Working with TeracPay, getting started and work, from Terac's own pages
What the platform publishes in general. These are not extra terms confirmed for this particular role.
- Getting paid
- Paid through Virtual Incentives: PayPal, Venmo, a prepaid Visa card or direct cashout in the US, PayPal elsewhere. A payout starts once a submission is approved, and cashing out needs a $12.50 balance and a verified ID. terac.com/docs/experts/policies/payout-policy, terac.com/docs/experts/how-to/payments, terac.com/whats-new/2026-05-31-may-2026-updates
- Before your first paid task
- A verified government ID through Persona before you can apply, then a screening step for each study: a short form, an AI voice interview or a file upload. A study you are screened out of is not paid. terac.com/docs/experts/how-to/apply, terac.com/docs/experts/policies/rewards
- How much work
- Terac guarantees no number of opportunities and no earnings. Manually reviewed applications hear back in 1 to 5 business days. terac.com/docs/experts/policies/terms, terac.com/docs/experts/support/faq
- Contract
- Independent contractor, with Terac Inc. of California. terac.com/legal/terms-of-service
Read on the pages linked above on 23 Sep 2026. Platforms change these terms, so check the linked page before you rely on one.
How does the pay compare?Across Cybersecurity roles, not this role alone
From our own daily reads of every open listing on the platforms we follow, not from Terac's posting. How we count.
Pay across the field: Cybersecurity
Across 59 roles in this field with a stated hourly rate in US dollars, on 11 platforms, the middle half pays $60 to $100 an hour at the top of the range, and the median is $75. This role's top rate, $150, is in the top quarter. Of the 32 distinct rates we count, 30 are lower and 2 are higher.
| Platform | Roles | Top of range, per hour |
|---|---|---|
| Alignerr | 32 | $55 to $120 · median $60 |
| Sepal AI | 5 | $60 to $100 · median $80 |
| Braintrust | 5 | $15 to $75 · median $45 |
| Mercor | 4 | $90 to $250 |
| DataAnnotation | 4 | $125 |
Sources and listing historyWhere this information comes from
This listing's history
- First seen here on 9 Nov 2026, and still on Terac's list when we last read it on 21 Nov 2026.
- The listed pay has not changed since then.
- Posted by Terac
- First seen by AIPayList
- Last seen live
- Listing ID
- SlJloh9LnnYmCp0U
Pay as the source states it: “Terac listing record, listed rate 150.00 USD HOURLY”.
Platforms sometimes change a rate after posting. Every listing here is re-read at least once a day and corrected when its source moves, so a difference you see on the platform is usually a change made since the last read. How freshness works.
Keep exploring
More like this
Related field, different requirements. Check each role's own page before you apply.