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.

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 (“LLMs”) upon which TaskUs develops its AI solutions are provided by reputable providers that are equally focused on AI compliance governance. As our capabilities expand, TaskUs is planning to develop proprietary AI and Agentic AI systems to offer advanced service solutions to our clients. 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. Moreover, TaskUs prioritizes data minimization especially if the AI tools process personally identifiable information (“PII”). Unless approved by exception, all PII data is redacted or anonymized before entering into an AI system. When PII processing is necessary for service delivery, it is done strictly within approved secured environments, and in accordance with client instructions. To confirm that AI tools integrated into its services are secure, TaskUs completes vendor security and data privacy assessments on the AI tools prior to considering them for use.

Client trust and data security are TaskUs’ highest priorities. TaskUs prohibits processing any client data or PII through any AI system without the explicit approval of our clients. While client data is not used for our internal AI initiatives, we do responsibly leverage AI systems and tools to help improve our foundational, non-customer-facing operations. These internal applications are restricted to areas such as workforce management, business intelligence, learning experience (training), and other critical areas foundational to the delivery of service to our clients, without exposing client-specific data.

1. Our use of AI Tools should be RELIABLE & FAIR. TaskUs endeavors that its 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. To maintain a culture of transparency, internal outputs significantly generated or altered by AI (such as automated reports, company-wide announcements, or training modules) should be clearly labeled as ‘AI-Assisted’ or ‘AI-Generated’ to manage expectations regarding accuracy and origin. Further, we implement mechanisms for meaningful human intervention where AI aids in decisions that significantly impact individuals allowing for recourse or manual review.
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. While TaskUs believes AI should augment rather than replace human judgment, it implements a human-in-the-loop model that is proportional to the risk and complexity of the task. AI-driven internal automations shall not be the sole basis for high-stakes employment decisions, including performance evaluations, disciplinary actions, or hiring. AI-generated internal recommendations that affect employment must be reviewed by a qualified human lead to ensure fairness and prevent algorithmic bias.
5. Our use of AI should be PRIVATE AND SECURE. AI systems generally process data based on the instructions of our clients. TaskUs ensures that any personal data processed by AI is done so in accordance with the lawful basis established by the client, Data Processing Agreement (DPA) and the established applicable laws.

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.

All personnel, from executive leadership to frontline teammates, are prohibited from utilizing unauthorized or non-enterprise AI tools for any company-related tasks. All internal automations must leverage TaskUs-approved platforms exclusively, with every specific use case undergoing vetting by the AI Technology, InfoSec, Core Apps, Legal, and Privacy teams, as the case may be, prior to production or general implementation. Furthermore, employees seeking to utilize high- or medium-risk AI tools may be required to execute additional accountability forms to formalize their commitment to responsible and ethical use..

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. Unless an exception is approved, we will implement Privacy Enhancing Technologies (PETs) such as PII redaction, anonymization, and data minimization techniques before data enters an AI system, ensuring only essential information is processed.

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 and Data Privacy Impact Assessment 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.

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.

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.