AUTOMATIC DATA PROCESSING INCADP

ADP offers ADP Assist AI Agents built into its HCM platform.

“• Lead with Best-in-Class HCM Technology. We design and develop world-class, AI-enabled HCM platforms that simplify work and help organizations manage their most critical workforce functions with precision, compliance and confidence in the AI era. Unlike generalized AI tools, our AI is built into the very core of how we orchestrate, govern, and execute HR and pay processes for real-world outcomes. Our global data platform, named “2026 Data Solution of the Year for HR” in the Data Breakthrough Awards, is the foundation of the advantage we bring to our clients. Spanning over 1.1 million clients and 42 million workers across roles, industries and geographies, we have the industry's largest workforce dataset. In the U.S., we pay one in six workers and moved $3.5 trillion in fiscal year 2026, giving us unique insights into the workforce and its emerging trends. AI is only as good as the data it is trained on, and this advantage compounds over time. Every ADP Assist AI Agent is grounded in ADP's institutional knowledge. The result is persona-based agents tailored for employees, managers, HR and payroll practitioners, all informed by ADP's data advantage.”

ADP offers purpose-built AI agents recognized for solving client challenges.

“Our approach to AI agents earned ADP recognition as one of Fast Company's 2026 “Most Innovative Companies,” with ADP earning the number one spot among HR companies for its purpose-built approach to designing AI tools that solve real client challenges and empower people at work.”

ADP invested in research and development including generative and agentic AI.

“During the fiscal years ended June 30, 2026, 2025 and 2024, we invested approximately $1.405 billion, $1.388 billion, and $1.276 billion, respectively, in research and development. These investments include expenses for activities such as the development of new products, maintenance expenses associated with our existing technologies, investments in generative and agentic AI, purchases of new software and software licenses, and additions to software resulting from business combinations.”

ADP is integrating generative and agentic AI but faces legislative and ethical risks.

“We believe that providing insights and content from data, including via AI and ML, will become increasingly important to the value that our solutions and services deliver to our clients. We are increasingly leveraging AI and ML in our solutions and service delivery and are continuing to integrate AI technologies, including generative and agentic AI, to develop and deploy capabilities that are beneficial to our clients and their employees. However, legislation that governs the development and/or use of AI has been adopted or is under consideration in the U.S. at the state and local level, as well as abroad, most notably the European Union’s Artificial Intelligence Act. In addition , self-regulatory frameworks like the National Institute of Standards and Technology AI Risk Management Framework are being promulgated and adherence to these may become an industry standard or a client expectation. As a result, the ability to provide data-driven insights and otherwise leverage AI and ML may be constrained by current or future laws (including product liability regimes), regulatory or self-regulatory requirements or ethical considerations, including our own published, guiding ethical principles regarding AI and ML, that could restrict or impose burdensome and costly requirements on our ability to leverage data and/or these technologies in innovative ways. Our use of AI, including generative and agentic AI, in our products and operations also introduces additional risks, including risk”

ADP notes risks to intellectual property from increased use of generative and agentic AI.

“Our ability to compete and our success depend, in part, upon our intellectual property. We rely on patent, copyright, trade secret and trademark laws, and confidentiality or license agreements with our employees, clients, vendors, partners and others to protect our intellectual property rights. We may need to devote significant resources, including cybersecurity resources, to monitoring our intellectual property rights. In addition, the steps we take to protect our intellectual property rights may be inadequate or ineffective, or may not provide us with a significant competitive advantage. Our intellectual property (including source code) could be wrongfully acquired as a result of a cyber-attack or other wrongful conduct by third parties or our personnel, or as a result of increased use of AI tools, including generative and agentic AI, by us or our vendors. Litigation brought to protect and enforce our intellectual property rights could be costly and time-consuming. Furthermore, our efforts to enforce our intellectual property rights may be met with defenses, counterclaims, and countersuits attacking the validity and enforceability of our intellectual property rights, which may be successful. In addition, use of AI tools may result in the release”

ADP states AI technologies including agentic AI are used by threat actors to enhance cyberattacks.

“We have programs and processes in place designed to prevent, detect and respond to data or cybersecurity incidents. However, as a result of the complexity of our operating environment, the period over which hardware and software has been acquired or other reasons, our programs and processes may not be sufficient or adequate or may fail to prevent, detect or respond to a cybersecurity incident or identify and/or remediate a security vulnerability in our operating environment. The techniques used to obtain unauthorized access, disable or degrade service, or sabotage systems change frequently and are increasingly more complex, sophisticated and effective (including due to the use of AI). As AI technologies, including generative and agentic AI, continue to evolve, threat actors are increasingly leveraging these technologies to enhance the sophistication, scale, speed and effectiveness of cyberattacks making them more difficult to detect and defend against. The adoption and deployment of AI technologies within our and our authorized third parties’ solutions, services, and systems may also introduce novel security, data governance and operational risks. In addition, new computing technologies, including quantum computing, new discoveries in the field of cryptography or other developments could result in a compromise or breach of the algorithms we or our authorized third parties use or have used to encrypt and protect data. We may fail to anticipate or detect these techniques and/or”

ADP Assist agents launched in January saving client time on payroll and HR.

“During the third quarter, we made meaningful progress on our Strategic Priorities. We continued to leverage our data advantages, domain expertise, and trusted brand to lead the HCM industry through its AI transformation. Our ADP Assist agents are applying advanced intelligence to real workforce challenges across payroll and HR, and since their launch in January, our clients have seen meaningful time savings for each payroll processed. ADP Lyric HCM is also saving time for our clients, with one company replacing a dozen disparate systems to achieve a leaner payroll operations model. Additionally, we expanded our AI ecosystem by launching a dedicated space within ADP Marketplace for our partner companies’ AI agents. On the service front, we scaled our GenAI capabilities across our service operations, transforming how our client-facing teams engage, serve, and support our clients across the full lifecycle. Finally, we remained focused on benefiting our clients through our global scale by delivering compliant HCM solutions, local expertise, and trusted relationships wherever they operate.”

ADP states agentic AI is the future of innovation in the HCM space.

“The AI landscape is evolving quickly. As we move from early automation and predictive analytics to AI-enabled workflows, agentic AI is the future of innovation in the HCM space. Our AI strategy is to deliver AI agents for all”

ADP plans to deliver AI agents for HCM roles for clients and internally.

“major HCM roles that our clients have. Our clients have employees performing different roles such as payroll, HR, benefits, compensation, and recruiting. Our goal is to deliver AI agents that continue to automate tedious, time heavy tasks and deliver value to these individuals to do more strategic activities. Similarly, we intend to deliver AI role-based agents to our internal associates such as sales, software engineering, testing, product management, service, and implementation.”

ADP presents a slide on the evolving AI landscape including agentic AI.

“THE AI LANDSCAPE IS EVOLVING FAST Time Business Value Agentic AI Reasoning and decision-making Source: Internal research *RPA: Robotic Process Automation; AI/ML: Artificial Intelligence / Machine Learning; NLP: Natural Language Processing; GenAI: Generative AI Early automation RPA* Data-driven insights Predictive Analytics (early stage of AI/ML*) Understanding language NLP* AI Workflow GenAI* + RPA* Simple Agent Advanced Agent Autonomous Agent 26”

ADP describes role-based AI agents for HCM tasks and journeys.

“AGENTIC AI IS THE FUTURE Role-based AI agents Payroll | HR | Benefits | Recruiting Customized Recommendations Execute multi-step HCM Journeys Anomaly detection Task Automation Interpret & Assess Information Monitor & Manage Interactions 27”

ADP highlights its data advantage for developing smarter AI agents.

“DATA AS A DIFFERENTIATOR ADP’s 1.1M+ clients and millions of transactions Training on broader, more granular data, delivers smarter and more tailored agents AI agents develop hire-to-retire expertise, equipping them to automatically execute complex HCM journeys ADP’s large product portfolio usage data, sales data, and operations data 28”

ADP embeds role-based AI agents into products and delivered ADP Assist.

“STRATEGICALLY INFUSING AI ACROSS THE BUSINESS Embed role-based AI agents into our products to improve client experience • Proactive Nudges • Recommended best practices • Automated task completion Delivered ADP Assist across several products; driving adoption STRATEGY 29 ~4M Interactions on ADP Assist1 1 FY25 year-to-date PROGRESSPRODUCTS1 2 SERVICE 3 SALES R&D4”

ADP plans to expand AI efforts into agentic solutions for small business.

“Achievements • Set the foundation for RUN’s cloud- native modernization journey • Tightly integrated Retirement & Insurance Services • Launched Embedded Payroll offering 60% of RUN clients digitally onboarded 1 Search enhanced by ADP Assist contributed to improved client self- service • Enhance HR experience for growth in mature small businesses • Scale Embedded Payroll • Expand digital buying experiences What’s Next Expand AI efforts into agentic solutions 1 Q3FY25 REIMAGINING SMALL BUSINESS RUN Powered by ADP® — 900K+ clients 34”

ADP launched ADP Assist and plans to expand AI into agentic solutions.

“Achievements • Continued momentum with WFN Next Gen sales • Expanded partner channels to cover key mid-market ERPs • Launched Construction industry vertical Launched ADP Assist with Payroll across 100% of WFN Next Gen clients 2M HR conversations across 1.3M unique users spanning 68k clients 1 Search enhanced by ADP Assist contributed to improved client self- service • Extend WFN Next Gen addressable market • Scale ERP partner channel • Expand into Restaurant vertical What’s Next Enhance AI-driven Tax and Payroll experiences Expand AI efforts into agentic solutions 35 INNOVATING TO REMAIN THE U.S. MID-MARKET LEADER Workforce Now® (WFN) — 90K+ clients 1 FY25 year-to-date”