Skip to content
Let's Meet
Let's Meet
    November 8, 2023

    AI Governance for HR with Guru Sethupathy: FairNow Product Profile

    I recently had the privilege of sitting down with Guru Sethupathy, the visionary behind FairNow, an AI governance platform designed specifically for human resources. With over 15 years of experience at the intersection of analytics, AI, and human capital, Guru is well-versed in the potential of AI and the importance of implementing responsible AI practices.

    FairNow AI Logo

    Meet Guru Sethupathy of FairNow

    Guru Sethupathy HeadshotGuru Sethupathy, Co-founder & CEO of FairNow

    Guru Sethupathy's journey into the world of AI and HR started during his time in academia, where he conducted research on the profound implications of AI on the workforce, jobs, and skillsets. He was keen on understanding how AI could revolutionize the talent industry, which, until recently, had been operating on age-old principles and methods.

    According to Sethupathy, the talent industry is stuck in a time warp, using the same principles and processes for decades. There was a significant opportunity for disruption, and we've been witnessing that disruption with the evolution of people analytics, HR technology, and AI entering the HR domain.

    Over the past decade, Sethupathy has had the privilege to work at prestigious institutions like McKinsey and Capital One, where he strategically advised and engineered digital experiences and machine learning solutions in highly regulated environments like HR for Fortune 100 companies. Throughout his journey, he has always adhered to a fundamental framework of responsible AI and governance.

    Responsible AI and Governance

    Sethupathy’s commitment to responsible AI stems from a deep belief in the technology's potential and the transformative power it holds. However, he also recognizes the risks associated with unchecked AI deployment.

    "If you don't build and deploy this technology in a trustworthy and responsible fashion, it is going to turn a lot of people off. I believe that it's going to cause regulatory and reputational burdens and challenges and risk, and it's going to lower people's desire and investment in the technology. And I think that's a loss for everyone.” - Guru Sethupathy, Co-Founder & CEO of FairNow

    This belief in the mission of helping organizations establish robust governance practices around their AI and demonstrating the trustworthiness of their AI systems led to the inception of FairNow. The platform's primary objective is to help companies automate Rigorous governance practices around AI and to showcase that their AI systems are both reliable and responsible.

     

    Addressing the Imperative of AI Governance in HR: The Need for a Product Like FairNow

    During our conversation, we delved into the necessity of AI governance in HR. One of the challenges that Sethupathy highlighted was the potential governance blind spots in organizations, particularly concerning data models. He explained that not all data models should be subject to governance, and sometimes, even the AI may not be aware of this. It's essential to identify which data sets or models require AI governance.

    Sethupathy approaches governance with a nuanced perspective. His first step is to conduct a risk assessment of each AI model, considering factors like the use case and usage frequency. "Use cases with significant impacts on individuals' careers, such as hiring decisions," he pointed out, "are considered high risk."

    The European Union has designated certain domains, including hiring, as high risk. On the other hand, use cases with minimal impact on individuals' careers are deemed low risk. Sethupathy emphasized that even low risk models may require governance. Inventorying, tracking, metadata assignment, and ownership identification are fundamental governance practices, even for low risk models.

    The key dimensions for a risk assessment include the use case's impact on individuals, usage frequency, and relevant regulations. FairNow allows employers to tailor their governance based on the assessed risk level. High risk models require more stringent governance, while low risk models would have a more streamlined governance framework.

     

    FairNow's Category and Maturation in AI Governance

    Guru acknowledges that while AI governance is relatively new, it shares common ground with well-established domains like Information Security (InfoSec). InfoSec, responsible for the management of information security, has been a staple in the corporate world for a considerable time. Companies have long adhered to standards like SOC 2 compliance, which serve as benchmarks to ensure the effective management of information security within an organization. “We're a startup, Drew, and we will not be able to get any customers if we're not SOC 2 compliant.' This is a compelling indicator of where responsible AI is heading," Sethupathy said.

    In fact, SOC 2 compliance is so essential that startups, even those just starting out, know that adhering to such standards is non-negotiable. These compliance standards are vital in building trust with stakeholders, be it the board of directors, customers, or partners.

    Now, as the landscape of responsible AI governance emerges, it is clearly expected to follow a similar trajectory but with even greater demands. For instance, Sethupathy highlighted the potential future necessity of being ISO 42001 certified in responsible AI to gain the trust and consideration of customers.

     

    How is FairNow - an AI Governance Software for HR, Changing the Game?

    FairNow offers a platform that addresses the complexities of AI governance in HR with three core capabilities:

    1. Managing governance programs

    FairNow streamlines the establishment of comprehensive governance programs by defining essential roles, developing AI policies, and offering frameworks for managing AI data models. This integrated approach helps organizations enhance accountability, transparency, and overall compliance with responsible AI standards.

    2. Tracking AI laws and regulations

    FairNow keeps organizations informed about AI-related laws and regulations. It tracks the rapidly legal landscape of AI and assesses whether organizations are compliant. For organizations falling short of compliance, FairNow provides guidance on necessary steps for alignment.

    3. Conducting automated bias audits for high-risk models

    Bias audits are integral to any comprehensive AI governance program. FairNow conducts automated bias audits for high risk models. These audits are conducted in multiple ways to ensure that AI systems are free from bias.

    FairNow’s Governance Program

    Sethpathy also gave us a demo of how FairNow’s governance program works. He highlighted three essential components of this program:

    AI Governance Roles

    FairNow requires every organization that registers to identify some key roles. These are:

    Risk Officer: Responsible for implementing model policy assessing AI-related risks within the organization.

    Accountable Executive: Sponsors the project and takes ownership of risks.

    Model Owner: Is usually the data scientist who built the model and has access to create and edit model reports.

    AI Governance Policies

    The next step is having AI policies in place. These include:

    Transparency Policy: Transparency is essential, but the right level of transparency needs to be defined. How much are you willing to share to which group of stakeholders?

    Accountability Policy: These define who is responsible for AI-related decisions and actions. They help ensure that AI processes are accountable and aligned with organizational goals. Who is responsible for approving a model? Even if a model makes a business decision, who is the human that is accountable for that decision?

    Fairness Policy: Fairness policies address the critical concern of bias and discrimination, promoting equal opportunities for all candidates and employees. Fairness can be defined in multiple ways and organizations will need to align on their definitions of fairness.

    Ethics Policy: All processes should adhere to ethical guidelines, covering areas like data privacy, consent, and ethical decision-making, and these policies help define those guidelines. Under what situations would you not consider using AI?

    Privacy Policy: Privacy policies are essential in the HR AI landscape to protect the sensitive data of employees and candidates.

     

    AI Governance Processes and Controls

    The last component is processes and controls. FairNow takes well-established frameworks, such as NIST, EUAI Act, and ISO, and translates them into practical actions that organizations can seamlessly implement. These actions encompass the entire AI model development lifecycle, including the building and monitoring phases.

    They empower organizations to methodically follow these steps, ensuring they adhere to responsible AI principles and best practices. As Guru pointed out, this approach not only results in a robust governance structure but also allows organizations to proudly attain certifications like ISO certification, which can be presented to their board and customers, further solidifying their commitment to ethical AI governance.

     

    Empowering Responsible AI: FairNow's Impact

    FairNow stands as a beacon of responsible AI in HR. With its governance capabilities, role definitions, and AI policy templates, it empowers organizations to build trustworthy and responsible AI systems with ease. By helping companies navigate the complex world of AI governance, FairNow is instrumental in establishing a culture of responsible AI use in HR.

    One of FairNow's standout features is its automated bias audits for high risk AI models. Bias in AI can lead to discriminatory outcomes and perpetuate existing inequalities.

    The platform leverages statistical methods and fairness metrics to assess AI models for bias. It also conducts external audits and comparisons to identify and rectify biases.

    For low risk models, FairNow recommends a more streamlined approach. It conducts risk assessments and uses guidelines to determine the extent of bias audits required.

    Why Trust FairNow

    FairNow's founders, with their background from Capital One, bring a wealth of experience in AI and data governance, making the company a trusted industry leader. Their years of involvement in the field have culminated in the development of FairNow, a platform that encapsulates best practices from their extensive experience.

    What sets FairNow apart is its commitment to cost-effective AI governance. Traditional bias audits can be a substantial expense, often exceeding $50,000 for a single tool. FairNow changes the game with its complete automation, making AI governance both efficient and accessible.

    The FairNow platform also offers a unique feature—continuous monitoring. Organizations can set up a regular cadence for running simulations, ensuring they are always audit-ready and fully compliant. FairNow's forward-thinking IP includes innovative simulation technology, solving common audit challenges. This creative approach accommodates companies without demographic data and those hesitant to share their data.

    FairNow is more than just a technology company; it offers white-glove service to its clients and additional service support through talent strategy advisors like the team at Lever Talent, Inc. While the platform is robust, the FairNow team understands that AI governance can be complex and new territory for many organizations. They are readily available to guide clients through the entire process, from defining roles and responsibilities to establishing their governance program.

    The consulting and expertise offered by FairNow are integral to a successful AI journey. The team provides valuable insights and ensures that clients comprehend and effectively implement AI governance.

    Steering HR's AI Transformation: The Role of Responsible AI Governance

    AI governance in HR is more critical than ever. It ensures that organizations benefit from the potential of AI while safeguarding against risks, biases, and unethical practices. FairNow, with its comprehensive AI governance platform, is a guiding light for organizations navigating the complex world of HR AI. Through defining roles, establishing policies, and automating bias audits, FairNow helps organizations build a culture of responsible AI use in HR.

    AI is transforming HR, and with platforms like FairNow, organizations can make this transformation while safeguarding their values and principles. Responsible AI in HR is not just a buzzword; it's a mission that ensures a brighter, fairer, and more ethical future for the workforce.

    Lever Talent is a strategic partner and reseller of FairNow, the leading AI Governance Platform for HR. Speak to someone about FairNow.

    Drew Fortin

    Drew is a people-first, values-driven leader with nearly 20 years of growth strategy and team-building experience across retail, marketing technology, local media, and HR tech. He spent 7 years at The Predictive Index, where he was Chief Growth Officer responsible for the company's strategy to build the world's first...

    More from the blog

    View All Posts