Last updated: March, 2026
As a developer and provider of Artificial Intelligence (AI) services, Bullhorn’s responsibility centers on the ethical design, secure development, and transparent documentation of our AI capabilities. Bullhorn’s dedication to developing trustworthy and safe AI is demonstrated by our voluntary signing of the EU Artificial Intelligence (AI) Pact.
Where model training is concerned, we carry this out in alignment with global data privacy laws, evolving AI regulations, and contractual obligations. We are committed to assisting our customers in their compliance efforts by providing detailed documentation and transparency into our AI services. Conversely, our customers, as the deployers, users, and, in many cases, employers, bear the responsibility for how they implement and utilize these AI services in compliance with the laws and regulations applicable to their specific use cases and hiring/employment decisions.
Our Core Principles for Responsible AI
Our commitment to responsible AI is built upon the following foundational principles, supported by industry best practices established by leading frameworks such as the NIST AI Risk Management Framework, the OECD AI Principles, and globally recognized AI governance standards:
- Fairness and Inclusivity: We strive to develop AI technologies that treat individuals fairly and do not contain systematic bias. We actively work to minimize unintended bias factors in our AI systems and set guidelines in order to limit misuse or exploitation. Our design choices exclude sensitive information such as gender, age, and ethnicity from search and ranking processes and we use query enrichment and objective ranking to reduce unintended bias.
- Reliability and Safety: Bullhorn’s AI systems are designed to perform reliably and safely. We implement robust security protocols and encryption to protect sensitive data consistent with industry standards. All AI systems undergo human review for accuracy, relevance, completeness, and compliance with our ethical guidelines and applicable laws.
- Privacy and Security: We develop AI capabilities in a manner that protects personal data and complies with privacy laws. Our policy strictly prohibits unauthorized data processing, especially of personal, sensitive, or confidential information. We adhere to data minimization principles, ensuring only necessary data is processed by the AI systems, and sensitive data is not used for matching/scoring.
- Transparency: We are committed to communicating about our product developments, policies, and processes where practicable and reasonable to maintain trust with customers. We aim for transparency in our AI tools, such as keyword-based matching algorithms that allow users to understand how results are generated.
- Accountability: We have established internal frameworks for oversight, insight, and guidance to facilitate responsible AI operations, ethical design and secure development of the AI systems we provide.
- Compliance with Applicable AI Laws: We stay abreast of industry standards and national, international, and local legal and regulatory requirements that apply to Bullhorn to ensure adherence to those standards and requirements.
How Bullhorn supports bias management and data oversight by customers within its AI systems:
- Mechanisms to Flag and Mitigate Bias: We integrate design choices to reduce potential bias, such as excluding personal attributes from search and ranking processes and expanding searches with synonyms to mitigate bias from overly specific keywords. Our matching processes are based on objective criteria like work experience and skills.
- Human Oversight: Our systems support decision-making but do not automate it. Users are required to review the results, and the ultimate decision rests with the user, based on their internal criteria, thereby ensuring human oversight.
- In alignment with evolving global AI regulations regarding transparency and fairness, we have partnered with an independent third party to conduct bias assessments of our core AI products. These audits scrutinize our feature selection process, differential impact across various demographic groups etc. The results were favorable, providing confirmation of Bullhorn’s commitment to mitigating algorithmic bias and ensuring equitable outcomes. For a complete understanding of the findings, the detailed reports are available upon request. Please reach out to your account manager(s) to obtain access.
- We support AI transparency and accountability through detailed, up-to-date technical documentation of our AI systems, which aligns with regulations such as the EU AI Act. This documentation covers:
- System Architecture
- Model Training & Validation
- Risk Mitigation
- Performance & Monitoring
- Compliance & Governance
Overall, Bullhorn is deeply committed to responsible AI, built on the core principles of fairness, reliability, privacy, transparency, accountability, and compliance. We actively support bias mitigation through design, ensure human oversight is built into the AI systems, and uphold transparency through documentation and third-party assessments. This commitment reinforces our dedication to developing ethical AI that empowers our customers while adhering to the highest standards of integrity in a continually evolving landscape.
Data & research
