Adversarial Prompt Expert
Adversarial Prompt Expert is an open role at Handshake AI, with pay listed as Up to $65/hr.
Applications open on Handshake AI’s own site. How we make money.
About this role
As an Adversarial Prompt expert, you'll be able to participate in project-based opportunities focused on probing large language models for failure modes and harmful outputs. Your work will involve crafting prompts and scenarios to test model guardrails, exploring creative ways to bypass restrictions, and systematically documenting outcomes. You’ll think like an adversary to uncover weaknesses, while collaborating with engineers and safety researchers to share findings and improve system defenses.
Pay
Up to $65/hr
About Handshake AI
Handshake AI's connection to Handshake: Handshake AI fellowship / opportunity network. Generalist and specialist AI evaluation and training projects
Ownership confidence: Confirmed. See the full network profile →
Similar open roles
Frequently asked questions
How do I apply for the Adversarial Prompt Expert role at Handshake AI?
Applications go through Handshake AI's own portal - HumanSourcer links to the listing and never collects applications or handles hiring. Handshake AI's access model is "Apply" (Apply to listed opportunities). This listing was first seen here on July 20, 2026 and was still live at the last check.
Is the Adversarial Prompt Expert role at Handshake AI remote?
Handshake AI left the location field empty on this listing, so there is nothing to report rather than something we've withheld. The constraint that still applies is Handshake AI's overall coverage: Many roles require US work authorisation.
What does the Adversarial Prompt Expert role at Handshake AI pay?
Handshake AI lists this role's pay as Up to $65/hr, quoted from the listing rather than estimated.
What kind of work is Adversarial Prompt Expert?
AI model evaluation. Generalist and specialist AI evaluation and training projects
Is Handshake AI a legitimate AI-training platform?
Handshake AI fellowship / opportunity network - and that relationship is rated "Confirmed" here because it can be checked against joinhandshake.com rather than taken on Handshake AI's word. Confidence describes how well the ownership is evidenced. It is not a rating of how the platform treats the people who work for it, which no public source covers reliably.
Get new AI-training roles by email
One weekly digest of newly listed roles across every tracked network - pay where it's disclosed, no spam, unsubscribe anytime.
Free. See how we make money.