TaskUs AI Global Policy
I. What is TaskUs’ Global Policy on the use of AI?

The TaskUs Global AI Policy (“Policy”) provides guidelines to its employees for the responsible use of AI tools at TaskUs, aiming to protect the company and minimize risks such as misuse, unethical outcomes, bias, inaccuracy, and security breaches. These guidelines aim to position TaskUs as a leader in AI-driven services by enhancing efficiency, innovation, and service excellence while complying with AI regulations and promoting the ethical and responsible use of AI in both client services and its internal operations.

For this Policy, a “client” refers to the businesses or organizations that partner with TaskUs to receive outsourced services, including AI-powered solutions, while “customer” refers to the end-users or consumers who interact with or benefit from the services TaskUs delivers on behalf of its clients. These can include the customers of TaskUs’ clients.

II. How Does TaskUs Use AI Tools in its Services?

TaskUs integrates AI tools into its services to enhance the speed, accuracy, and efficiency of service delivery to its clients while ensuring that humans remain at the core of critical decision-making processes. The Large Language Models (“LLM’s”) upon which TaskUs develops its AI solutions are provided by reputable providers that are equally focused on AI compliance governance. While AI tools are used to support operations, data processed by AI are subject to human review, and final decisions are made by humans to maintain accountability and accuracy. TaskUs does not develop standalone LLM generative AI models or systems. Instead, TaskUs uses trusted third-party AI tools and responsibly develops solutions to build and enhance services to clients. Further, TaskUs avoids using AI to process personally identifiable information (PII), prioritizing data privacy and regulatory compliance. To confirm that AI tools integrated into its services are secure, TaskUs completes vendor security assessments on the AI tools prior to considering them for use.

III. What are TaskUs’ guiding principles for using AI in its services to ensure compliance?
  1. Our use of AI Tools should be RELIABLE & FAIR. TaskUs ensures that AI systems and tools are reliable, accurate, and meet business needs by leveraging reputable third-party providers and rigorously testing their AI models. TaskUs developers will develop AI systems or tools with a goal of ensuring accurate results. TaskUs also commits to mitigating bias and regularly verifying that vendors continuously improve AI models to ensure validity, reliability, and fairness.
  2. Our use of AI should be TRANSPARENT. The processes and decisions made or aided by AI systems should be clear, understandable, and explainable to help users understand how a specific decision or output is arrived at or aided by the AI.
  3. Our use of AI should be DYNAMIC. AI compliance involves the continuous improvement and adaptation of AI systems to meet evolving ethical standards and technological advancements and we in turn must ensure that AI systems remain effective, relevant, and aligned with ethical principles over time.
  4. Our use of AI should provide ACCOUNTABILITY. There should be a way to hold those responsible for developing and using AI systems accountable for their actions to ensure that AI systems are used ethically and responsibly.
  5. Our use of AI should be PRIVATE AND SECURE. AI systems should respect the privacy of users by not collecting or using personal data without the user’s consent to protect the paramount right of individuals to their privacy and data.
IV. In what levels of the organization does TaskUs ensure responsible use of AI?

TaskUs requires responsible use of AI at all levels of the organization. Each team, including Marketing, Sales and Customer Service, Digital, Information Security, Core Applications, Data Privacy, and Operations follows applicable guidelines to comply with legal regulations and industry best practices. Further, all AI tools procured from third party vendors must be fully vetted and approved by TaskUs and/or the client prior to use.

V. How does TaskUs ensure data privacy while using and integrating AI tools?

TaskUs is dedicated to prioritizing data privacy. We utilize data processing agreements to ensure that all data processing activities comply with relevant laws and strictly adhere to client instructions. These agreements outline our obligations and responsibilities for handling, processing, and protecting data, ensuring that all actions meet both the client’s data privacy requirements and legal standards.

VI. How does TaskUs choose AI vendors?

TaskUs prioritizes selecting vendors with a proven and reputable track record in AI development. The Procurement Team conducts a comprehensive evaluation of the vendor’s history, including client testimonials, adherence to industry standards, and possession of relevant ISO certifications. Additionally, every AI vendor undergoes a Vendor Security Assessment by TaskUs’ InfoSec Compliance Team before their AI tools are approved for use. This ensures that all tools meet the industry standards of security and compliance before integration into TaskUs operations.

VII. How will TaskUs train its employees to adhere to responsible use of AI?

TaskUs recognizes that an informed workforce is the best line of defense. TaskUs provides training and expert resources to help employees understand their obligations under this Policy and appropriate use of AI tools. TaskUs may deem failure to participate in required training a violation of this Policy. TaskUs will retain training records and copies of generative AI training materials provided to employees.

VIII. What should clients do if they have concerns about AI usage?

TaskUs is committed to establishing clear lines of responsibility, along with effective mechanisms for reporting and addressing grievances, to ensure full compliance with ethical guidelines and regulations.

Clients can reach out to their assigned TaskUs Point of Contact for any concerns or questions regarding the use of AI in our services. For reporting violations of this Policy or any legal regulations, an email to [email protected] may be sent. We are committed to addressing these concerns promptly and ensuring that our AI systems meet the highest standards of ethics and compliance.

IX. How is this Policy Administered?

The AI Advisory Board, composed of the different team leads, provides guidance on the administration of the Policy. Key responsibilities include advising on training resources for the compliant use of AI for content creation, recommending regular audits of AI tools in partnership with relevant departments, and consulting with the Legal Department on any legal concerns. Meanwhile, the Vendor Management Office (VMO) acts in an oversight capacity. Through collaborative efforts, the AI Advisory Board and the various assigned groups help ensure that AI tools are used ethically and in accordance with the Policy, with accurate records maintained and issues escalated to upper management when necessary.

This Policy will be reviewed annually starting on October 5th of each year, to ensure its relevance and effectiveness in supporting the TaskUs’ AI governance goals. Amendments may be made to reflect changes in organizational priorities, regulatory requirements, or technological advancements in AI.