Platform Safety Engineer: Build Defenses Against AI Misuse

Infinized Cookie settings PeaceWeb Group; Nederland; Hybrid; EUR 75,000 - 110,000; Full time

Job summary

Infinized is seeking an engineer to build safeguards that stop misuse before it happens, covering backend systems, abuse and fraud prevention, and LLM defenses. You will design and implement protections across runs, prompts, and communications, while ensuring privacy and safe defaults across Infinized Voice and related systems.

We value attackers’ mindset, experience with production systems, and a pragmatic approach to security, with strong emphasis on preventing false positives and maintaining

Qualifications

• 5+ years building backend systems in production, including abuse or security problems.
• Ability to think like an attacker and build defenses that keep systems running.

Responsibilities

• Detect misuse in runs by identifying signals in run histories that indicate spam, fraud or deceptive AI agents.
• Defences against prompt injection to prevent new orders being given to AI agents.
• Set sensible limits on outgoing mail and messages to prevent mass misuse.
• Ensure safety for Infinized Voice with fraud checks and consent rules before going live.
• Provide analysts with a unified view and tools to understand cases quickly.
• Apply defaults that keep new AI agents on the safe side so safety isn’t dependent on every setting.

Skills

Backend systems

Security mindset

Abuse and fraud

LLM defenses

Security and privacy

Job description

Infinized is seeking an engineer to build safeguards that stop misuse before it happens, covering backend systems, abuse and fraud prevention, and LLM defenses. You will design and implement protections across runs, prompts, and communications, while ensuring privacy and safe defaults across Infinized Voice and related systems.

We value attackers’ mindset, experience with production systems, and a pragmatic approach to security, with strong emphasis on preventing false positives and maintaining