About the Company

• Dar, the founding member of the Sidara group, is an international multidisciplinary consulting organization specializing in engineering, architecture, planning, environment, project management, facilities management, and economics.
• Sidara operates in 60 countries with 20,500 professionals, Dar connects people, places, and communities through innovative solutions to the world's most complex challenges.
• We deliver projects from inception through completion, embracing challenges to empower communities worldwide.
• Learn more at www.dar.com.

Our Vision and Values:

• We aspire to be the chosen home of those with a gift for crafting solutions that empower people and an unwavering passion for learning and innovation.
• Our core values shape our culture and guide our decision-making.
• We are committed to:
• Excellence
• Responsibility
• Empowerment
• Connectivity
• Courage

Job Purpose

The role of the AI Security Engineer at Dar will be to secure the artificial intelligence systems used across the organization by designing and implementing guardrails, applying the security controls defined by the Cybersecurity team, and safeguarding the models, APIs, data pipelines, and applications that support design and project delivery. Working in close partnership with Cybersecurity, AI engineers, and data scientists, the role also supports AI Technology leads in developing the procedures and standards that govern the responsible and compliant use of AI at enterprise level.

Key Responsibilities

• Design, implement, and maintain guardrails for existing AI and LLM-based systems, including input/output filtering, prompt-injection and jailbreak defenses, PII and sensitive-content redaction, tool and agent permission boundaries, and usage/rate limits.
• Implement the security controls identified and communicated by the Cybersecurity team across AI platforms, models, APIs, and data pipelines, covering identity and access management, secrets management, encryption, logging, and network segregation, and evidence that those controls are operating effectively.
• Support Technology and Information Security leads in developing enterprise AI security procedures, standards, and acceptable-use guidance
• Perform security testing of AI models and applications, covering prompt injection, insecure output handling, data exfiltration, and model supply-chain risks; document findings and track remediation with the owning engineering teams.
• Secure the AI/ML supply chain and MLOps pipeline through model and dataset provenance checks, dependency and container scanning, artifact integrity, and secure CI/CD for model deployment.
• Build alerting techniques for deployed AI systems to notify on AI misuse, guardrail bypass attempts, anomalous usage, and security-relevant model drift, escalating to the Cybersecurity team in line with agreed procedures.
• Collaborate with AI, data and security teams to develop data leak prevention service.
• Align with cybersecurity to integrate a suitable AI security assessment tool — to continuously evaluate the AI applications against security checks, rather than one-off manual reviews
• Assist Cybersecurity team in assessing the security and data-protection posture of third-party AI models, vendors, and APIs prior to onboarding, and document residual risks together with the required compensating controls.
• Maintain technical documentation, control records, and audit evidence to support internal reviews, client security requirements, and regulatory compliance.
• Stay current with the evolving AI threat landscape, regulations, and tooling, and contribute to internal awareness, mentoring, and knowledge sharing on secure AI practice.

Qualifications

• Must have a degree in Computer Science, Cybersecurity, Artificial Intelligence, Data Science, or related field
• Must have 1-3 years professional experience spanning AI/ML and information security, with hands-on exposure to at least two of the following: LLM or agentic application development, application security, cloud security, security engineering/DevSecOps, or machine learning model deployment
• Working knowledge of AI-specific risks and mitigations, such as the OWASP Top 10 for LLM Applications, MITRE ATLAS, and adversarial machine learning techniques, is preferred
• Must have practical exposure to guardrail and AI safety tooling (for example NeMo Guardrails, Guardrails AI, Llama Guard, Azure AI Content Safety, or equivalent), including evaluation and tuning of filtering policies
• Must have proficiency in Python and SQL; knowledge of C#/C/C++ is a plus
• Must be familiar with AI frameworks such as TensorFlow, PyTorch, or Hugging Face, and must have experience building or securing APIs for custom, cloud-based or third-party AI models
• Understanding of cloud and platform security fundamentals on Azure and/or AWS, including IAM, secrets management, encryption, network controls, logging, and secure containerized deployment, is preferred
• Familiarity with security and AI governance frameworks such as ISO/IEC 27001, ISO/IEC 42001, the NIST AI Risk Management Framework, and data protection principles under GDPR or equivalent regulations, is preferred
• Experience working with relational and/or non-relational databases and with the secure handling of personal and confidential data, is a must
• Must have excellent problem-solving and communication skills, with the ability to translate security requirements into practical engineering changes and to work effectively in a team environment
• Strong interest in AI systems security is a must
• Relevant certifications (such as CompTIA Security+, Azure/AWS security associate, CEH, or an AI security credential) are a plus.

Equal Opportunity Statement

• While we carefully review all applications, only candidates who meet the specified requirements will be contacted for further consideration, We appreciate your understanding and thank you for your interest.