We believe talent is the most important differentiator for any organization, and we are committed to finding new ways to help our clients attract and hire better talent, faster. We have seen the positive impact artificial intelligence and data can make on talent acquisition processes. Ethical concerns around the use and capabilities of talent acquisition technology are growing as fast as the rate of change in technology.
Moving forward, technology will enhance human productivity by extending the capabilities of people and automating routine work. The pace of technological change will outpace the ability of governments to regulate it, so it’s essential that forward-thinking businesses take a leadership position to define how we should and shouldn’t use technology ethically. Therefore, we believe it is important to outline our technology ethics. By setting an ethical foundation, we can continue to grow and advance in a high-tech world without sacrificing our standards or integrity.
Our ethics pledge reflects our commitment to ensuring that technology is used in a way that not only drives efficiency, but also promotes fairness and enhances the experience of stakeholders. We will continue to work closely with our clients to understand their unique situations and offer processes or techniques that work best for their organization. We will strive to gain alignment with our clients to ensure that the solutions we offer ultimately support our collective ethics.
We firmly believe that no technology or method should be used in the hiring process of our clients without being researched and evaluated. We pledge to only use technology that has been thoroughly vetted and in manners that comply with the following principles:
We strive to ensure that candidates understand the significance, purpose and privacy of every step in their hiring process. We will design our products to promote transparency and explainability between our clients and their candidates, including with respect to how candidates’ personal data is used. We will be clear and honest with clients about the technologies we use on their behalf. We will not use “black box” AI that does not meet professional standards set forth by international organizations such as the Organisation for Economic Co-Operation and Development (OECD).
We will evaluate every feature to ensure it is designed to exclude bias and limit scalability of existing biases in talent acquisition processes. We will only offer products that are designed to facilitate fair, measurable and legal hiring.
We believe that AI systems in talent acquisition should be robust, secure and safe throughout their entire lifecycle so that in conditions of normal use, foreseeable use or misuse, or other adverse conditions, they function appropriately and do not pose unreasonable risk. We believe AI tools should evidence robust security and a systematic risk management approach that includes privacy, bias and digital security. We design our products with privacy, bias and digital security in mind.
We believe that all parties participating in projects related to AI should be accountable for the proper functioning of AI systems and for the respect of the above principles, based on their roles, the context, and consistent with best practice.
When a job candidate encounters an assessment used in the selection process, it needs to make sense to them. Assessments can be an important part of the candidate experience, but they must be introduced in a clear, respectful fashion. From the test-taker’s perspective, content must appear to be relevant to the job, perceived as neither invasive nor confusing. Well-designed assessments can enhance the experience of applying for a job and promote heightened self-awareness. We believe that efforts should be made to minimize the time required to complete an assessment, and that the test-taking experience should be engaging and comfortable.
We believe in cultivating a culture of respect for job-seekers. When we ask an individual to spend time taking an assessment, we want that investment of time and attention to provide real value. The first step is building assessments that are grounded in behavioral science. From there, test development must reflect a careful and well-reported process. Our assessment products are backed by technical documentation that includes the scientific framework underlying our approach and a report of the testing used to establish the psychometric properties of the assessment.
Among these psychometric properties are an assessment’s validity and its reliability. The former speaks to a foundational principle: an assessment used in a hiring context must only measure material that is relevant to the job under consideration. If assessment results have no demonstrated relationship to a person’s on-the-job experience, then their use cannot be justified. We advise clients on the appropriate use of our assessments, including the proper pairing of assessments with particular job roles. To that point, some assessments are not appropriate as a means of evaluating candidates for certain jobs. When multiple assessments are deployed to evaluate a single candidate, we seek to ensure that end users understand the contrasting nature of resulting data and appreciate best practices for comparing candidates.
We believe that assessment data is best used in the context of a multi-pronged consideration of an individual, not as a fixed exclusion criterion. We facilitate this approach by providing content to inform users and contextualize assessment results. Relatedly, we implement a program of training and counsel for our clients to advocate for proper use, including monitoring the ongoing utility, of our assessment(s). We promote the responsible use of assessment data according to US and international standards and implement appropriate security standards for our own data handling protocols.
Finally, our assessment products adhere to the EEOC’s Uniform Guidelines on Employee Selection Procedures. We carefully screen for evidence of unjustifiable adverse impact arising from the use of our assessments in candidate selection, and our assessment development process includes investigating results for different demographic groups, as reported in our technical manuals. We have also implemented additional measures to promote fair use, such as a full-length practice test available to candidates prior to commencing certain assessments and easy-to-use results summaries of all assessments to educate both candidates themselves and hiring personnel.
Alongside our commitment to ethics and transparency, we also serve our clients by keeping tabs on regulation and legislation forming around the use of technology in the hiring process. As part of that commitment, Paradox proactively conducts formal, independent audits of our technology — both for our own knowledge and self-improvement and for our clients. Our most recent bias audit was completed on June 5, 2023 and was commissioned from Holistic AI, an industry leader in AI audits, in connection with New York City Local Law 144. The summary of each independent audit can be found below.