NASDAQ: KDK

The Kodiak Driver is not just an idea—it is operating without a human driver today. As of December 31, 2025, Kodiak Driver-powered vehicles have logged over 10,700 Cumulative Hours of Paid Driverless Operations.

We built the Kodiak Driver with key technological advantages and differentiators that include: (i) our customer-focused design, engineered for maintainability and uptime; (ii) cutting-edge AI-powered software; (iii) independence from the high-definition (“HD”) maps common in the AV industry; (iv) a common technology platform that is built for scale and adaptable to nearly any modern ground vehicle; and (v) our approach to modular hardware and remote assistance. Notably, we have accomplished this in a highly capital-efficient manner.

We serve customers in the long-haul trucking, industrial trucking, and defense industries. In December 2024, we believe we achieved a historic milestone by becoming the first company to deploy customer-owned and -operated driverless trucks in commercial service. We delivered these driverless trucks to Atlas Energy Solutions (“Atlas”), an oil and gas logistics provider which operates these trucks in the oil-rich Permian Basin of West Texas and Eastern New Mexico. Following successful real-world operations and the achievement of key performance milestones, in March 2025, Atlas committed to deploying the Kodiak Driver on 100 Atlas-owned trucks, subject to the terms of the Atlas MSA. Additionally, we work with some of the largest fleets in the United States, including J.B. Hunt, Werner Enterprises, C.R. England and Martin Brower. As of December 31, 2025, our customers have utilized Kodiak-owned autonomous trucks to deliver more than 12,600 revenue-generating loads across the southern United States.

The Kodiak Driver is also being utilized in the defense sector, where we believe it can support national security initiatives and critical government applications. To date, we have recognized in aggregate approximately $30 million in revenue under contracts with the U.S. military to adapt the Kodiak Driver for military vehicles. In early 2026, we were awarded a contract by the U.S. Marine Corps to integrate the Kodiak Driver into the Marine Corps’ Remotely Operated Ground Unit for Expeditionary Fires (ROGUE-Fires) carrier ground vehicle.

We expect to continue to grow our Driver-as-a-Service (“DaaS”) business model, which we launched in December 2024 in connection with our partnership with Atlas. Under our DaaS model, our customers are provided with access to the Kodiak Driver on customer-owned and -operated vehicles. Under this model, we generate revenue through either a per-vehicle or per-mile license fee. This flexible approach to pricing is designed to align with our customers’ diverse operational models, while generating predictable, recurring revenue for us. By integrating the Kodiak Driver into customer-owned fleets, we expect to build an asset-light business that can scale with our customers’ growth. We first implemented the DaaS model with Atlas in its industrial domain. Currently, we charge our long-haul customers a traditional per-mile or per-load fee to deliver freight on Kodiak-owned autonomous trucks. We plan to transition those customers to our DaaS model once we commence long-haul driverless operations.

Kodiak was founded by autonomy industry pioneer Don Burnette in 2018. Kodiak’s management team has decades of collective experience across AI, robotics, and AV technology. Our team of 342 employees, as of December 31, 2025, brings experience not only in technology but also operations, trucking and defense.

The AV Industry

For decades, AVs remained a fixture of science fiction. Progress was limited until the early 2000s, when the Defense Advanced Research Projects Agency (“DARPA”) catalyzed the field through a series of landmark competitions known as the DARPA Grand Challenges.

The DARPA Grand Challenges led to significant advancements in state-of-the-art sensors, compute and drive-by-wire systems necessary to support AVs. Further, breakthroughs in physical AI and computer vision, powered by increasingly sophisticated graphics processing units (“GPUs”), have accelerated autonomy from the laboratory to real-world deployment. In the United States, policymakers, witnessing these advances, have cleared a regulatory path for AV deployment in 24 states as of December 31, 2025.

AV technology is no longer a technology of the future – it is currently being deployed across several economically-critical sectors including trucking, personal transportation, last-mile goods delivery and defense, with the public increasingly embracing driverless vehicles. The autonomous trucking sector in particular has recently seen increased customer interest, as the COVID-19 pandemic highlighted the fragility of supply chains to consumers, policymakers and investors.

Our Competitive Advantages

The Kodiak Driver’s technological advantages and unique architecture reflects the experience and expertise of our team. We made key architectural decisions early in our development process enabling us to build robust software and modular, durable hardware in a capital efficient manner. Our competitive advantages include:

1.Customer-Focused Design

2.Cutting-Edge Physical AI Powering a Unified Virtual Driver

3.Independence from HD Maps

4.Common Technology Platform

5.Modular Hardware

6.Assisted Autonomy

7.AI Safety Agent

Customer-Focused Design

We designed the Kodiak Driver with our customers in mind, leveraging our team’s experience deploying and operating real-world products. Specifically, we focused on the factors that we believe are critical for our customers, such as ease of maintenance, reliability and smooth integration into our customers tools and operations. Our iterative and efficient autonomy development process allows our team to move with agility and speed to address our customers’ needs.

Cutting-Edge Physical AI Powering a Unified Virtual Driver

Our cutting-edge virtual driver incorporates advancements in physical AI and a fundamentals-based sensor-fusion system. The Kodiak Driver can operate in various complex environments such as highways, dirt roads and off-road, using Kodiak’s Modular Cognitive Architecture to understand the scene. Our neural net framework emphasizes redundancy, end-to-end learnability, interpretability, generalizability and cross-sensor learning. The flexibility of the Kodiak Driver is based on this data-driven system architecture. Our approach is both modular and verifiable, and it utilizes emerging physical AI techniques, including Vision-Language Models (“VLMs”), to easily integrate diverse data streams. Key to our physical AI is our iterative autonomy development process, which involves a defined testing approach and strong systems engineering, thorough yet cost-efficient validation, advanced simulation utilizing technology currently provided by Applied Intuition, Inc. and rigorous software system design. The AI models in our proprietary perception and planning systems are trained and continually improved based on the data we collect through multiple sensing modalities, including camera, radar, and LiDAR sensor returns while a Kodiak Driver-powered vehicle is operating, as well as data from an Inertial Measurement Unit, which is incorporated into the Kodiak Driver. Our supervised learning is based on both manually-processed data processed and AI pre-labeling modules. For example, we use publicly available vision language encoders that are available for commercial use. We train our models on GPU cluster infrastructure, either at our own on-premise data center or in cloud-based clusters.

We do not use any third-party data sources in our autonomous system—all of our datasets are proprietary and originate from data we collect while a Kodiak Driver-powered vehicle is operating. This includes data we collect on our own Kodiak Driver-powered trucks, as well as data we collect while operating our DaaS model on customer-owned trucks.

Independence from HD Maps

The Kodiak Driver operates without relying on HD maps that are common in the AV industry. Instead, the Kodiak Driver is designed to see the road, analyze its surroundings, use limited pre-existing information and incorporate real-time perception. We believe this approach better enables the Kodiak Driver to adapt to construction, obstacles, and shifting lanes.

Additionally, we believe this approach enhances the Kodiak Driver’s ability to operate in unstructured environments, such as the Permian Basin and military theaters, where maps are difficult, if not impossible, to build and maintain. We believe our independence from HD maps will enable broader scalability and resilience in complex and dynamic environments, giving our trucks the flexibility to navigate across a range of situations and locations.

Common Technology Platform

The Kodiak Driver is a single, common technology platform that is designed to operate across multiple vehicle types and in a variety of driving domains. We have already demonstrated the Kodiak Driver on a number of vehicle types, including Class 8 trucks, Ford F-150s and a Textron RIPSAW M3 treaded military vehicle. In addition, this common technology platform allows us to leverage learnings from one domain to the next. Our strategic decisions to expand the Kodiak Driver’s operating domain to highways, surface streets and unimproved roads has created a robust solution that allows us to leverage learnings across driving environments, creating a virtuous technology development cycle.

Modular Hardware

We recognize that truck uptime and efficient maintenance are important for our customers. We designed modular hardware that is easy to service and maintain without the need for specialized technicians. This technology enables our customers to keep their Kodiak Driver-powered trucks up and running. Our focus on meeting these needs is evident in the design of our SensorPods, which are engineered to be replaced in less time than it takes to change a tire. We also recently announced a strategic agreement with Bosch, a leading global supplier of technology and services, to collaborate on and scale the manufacturing of a production-grade, redundant autonomous platform. This platform contains the specialized hardware, firmware, and software interfaces that enable the Kodiak Driver to automate trucks—either on a vehicle production line or through an upfitting partner.

Assisted Autonomy

The Kodiak Driver is designed to operate across a variety of environments. We have enhanced this capability by developing an end-to-end service capability, which we call Assisted Autonomy. Assisted Autonomy blends remote support

with onboard autonomy to create a solution that enables flexibility and safety. When operating in Assisted Autonomy mode, a driver holding a commercial driver’s license sits at a remote assistance station with intuitive controls including a steering wheel, pedals and brakes to provide guidance to the Kodiak Driver. Assisted Autonomy integrates into the Kodiak Driver’s AI safeguards, and uses redundant, low latency communications to enable safe, high-reliability remote assistance. We developed our Assisted Autonomy system to offer reliable end-to-end solutions, leading to reduced development costs and additional operational flexibility for our customers, and operate our Assisted Autonomy system in partnership with Vay Technology Inc. (“Vay”). Vay’s stations and software tools enable communication with the trucks using multiple redundant communications protocols, offering a more flexible driver-out solution that facilitates broad operational capabilities.

Connectivity for Kodiak's Assisted Autonomy technology is powered by Verizon. Verizon's 5G and LTE networks allow Kodiak’s driverless vehicles to communicate with Assisted Autonomy drivers and send mission-critical communication between vehicles and our 24/7 Operations Center, with low latency over long distances in remote environments.

AI Safety Agent

Kodiak's perception system utilizes our proprietary AI Safety Agent that leverages generative AI-based VLMs to identify and address novel, complex or rare edge case scenarios that can be a challenge for more traditional perception techniques. Using this new approach, the Kodiak Driver can generally identify scenarios like dust devils, flooded roads and car fires. We believe this technology allows us to more efficiently handle the long tail of complex edge cases as we move towards long-haul driverless deployment.

Our Safety Approach

Safety is the foundation of everything we build. It is the driving force behind what we do and why we do it.

Building a safe AV is not about using a single approach or achieving a single metric. It is about being comprehensive – identifying reasonably foreseeable risks and building layers of systems and processes to mitigate those risks. This holistic, comprehensive framework, common in safety-critical industries, is called a safety case. We utilize our safety case to provide a structured, evidence-backed argument for why the Kodiak Driver is acceptably safe.

Our safety case enables us to foster confidence among customers, regulators and the general public. To complete our driverless safety case for our initial industrial trucking operating domain, we conducted a comprehensive risk evaluation spanning functional safety, behavioral safety, operational safety, cybersecurity and other areas of focus. The evidence from this effort was then assembled into our safety case structure to support the claim that our system is acceptably safe for driverless operations in our off-highway operating domain. We are now working to extend our safety case to additional operating environments, including long-haul trucking. We intend to complete our long-haul safety case, and commence long-haul driverless operations, by the end of 2026.

Our Safety Case Pillars

Core to our safety approach is the Probabilistic Risk Assessment (“PRA”), a technique pioneered by the nuclear and aerospace industries, to quantify behavioral safety risks. The PRA provides a rigorous, data-driven assessment of risk, supporting a measurable and evidenced-backed path to deployment readiness, while helping to identify the highest-priority areas for development.

The PRA breaks down risks into testable and measurable scenarios, starting broad and progressively narrowing into specific, quantifiable events like “static vehicle-in-lane exposures per mile.” This structured risk assessment enables a precise evaluation of our autonomy system’s performance under real-world conditions. By pinpointing key sources of risk, the PRA allows us to prioritize development towards the most critical scenarios. As we continuously expand and refine our

PRA model, we systematically determine new engineering priorities, ensuring continuous refinement and ongoing safety improvements.

In addition to our safety case framework, we work both individually and with our suppliers to ensure we comply with all applicable Federal Motor Vehicle Safety Standards (“FMVSSs”) and Federal Motor Carrier Safety Regulations (“FMCSRs”). Furthermore, we have established a strong set of principles, policies and procedures designed to promote safety throughout our organization. We work closely with both federal and state regulators to ensure that they are well informed of our operations and approach and commitment to safety.

Our Solution

The Kodiak Driver

We are focused on delivering a purpose-built, physical AI-powered ground autonomy solution that enables reliable and efficient driverless movement in a wide variety of environments. The Kodiak Driver’s software stack employs a physical AI-driven approach to ground autonomy, utilizing a single virtual driver across multiple environments, while the Kodiak Driver’s modular hardware is designed for easy maintenance with minimal training. The Kodiak Driver has demonstrated performance in complex environments with different trailer types and weights.

Oversight and Integration Tools

Kodiak has built oversight and integration tools that allow us to deliver a comprehensive and turnkey AV solution to our customers. This includes Kodiak OnTime, Kodiak’s proprietary suite of support services and tools designed to provide smooth integration with customer operations. Kodiak OnTime integrates the Kodiak Driver with our customers’ management systems, including fleet management, transportation management and yard operations. In addition, Kodiak OnTime enables real-time tracking and status updates, and includes in-field applications for inspections as well as services to keep trucks moving safely and efficiently. Kodiak’s 24/7 Operations Center gives our operations specialists real-time visibility into every Kodiak Driver-powered truck, supporting efficient and reliable operations.

A Business Model Tailored to Customer Operations

Kodiak offers the Kodiak Driver for a recurring license fee. Integrating the Kodiak Driver into customer-owned vehicles allows us to maintain a more asset-light business model. Under our DaaS model, which we launched in December 2024 in connection with our partnership with Atlas, we charge a per-vehicle or per-mile license fee, depending on which fee structure better fits into our customers’ operations. These license fees cover the use of the Kodiak Driver hardware and software platforms, access to software and routing updates, maintenance of the autonomous system, remote monitoring and assistance and access to Kodiak OnTime and other services.

We believe that integration with our customers’ information management systems through Kodiak OnTime will aid in customer retention, as our customers adapt their processes and tools specifically to incorporate the Kodiak Driver. We anticipate entering into long-term master service contracts with our customers which will provide for per-truck or per-mile license fees and terms of three to four years in our trucking business.

We will deliver certain components of our offerings, such as maintenance, hardware financing and insurance, in conjunction with our third-party partner network. We believe that this business model will allow us to scale while maintaining high margins – we succeed as our customers succeed.

Technology

Kodiak’s Self-Driving Software Stack

The Kodiak Driver uses a Modular Cognitive Architecture consisting of several parallel deep neural nets that simultaneously process large amounts of sensor data to perceive and understand the truck’s operating environment and its place within it, classify the objects and actors in its field of view, predict the behavior of those actors, and create and execute a motion plan for where to drive. Our data-driven AI framework emphasizes redundancy, end-to-end learnability, interpretability, generalizability and cross-sensor learning.

The Kodiak Driver pulls in information from its suite of sensors: cameras, radars and LiDARs. Each of these sensors has its own strengths and benefits, and we purposefully arrange them with overlapping fields of view designed to improve safety. The Kodiak Driver’s AI-powered perception system then processes that sensor information, turning raw sensor data into actionable information. Next, the Kodiak Driver uses this information to determine its location, orientation and speed in relation to other objects and actors. It then formulates a motion plan for where to drive, setting both speed and trajectory to establish a safe path. Finally, the Kodiak Driver’s controls layer communicates with the underlying truck, applying the throttle or brakes and turning the wheels, as applicable. This full cycle repeats itself ten times every second.

The Kodiak Driver also utilizes technologies that promote safe and reliable operations even in challenging situations. Kodiak’s proprietary fallback system is designed to enable a Kodiak Driver-powered vehicle to safely and autonomously move over to the side of the road in the event of a truck or system failure, thereby achieving a minimal risk condition. Ten times each second, the Kodiak Driver evaluates the performance of more than 1,000 safety-critical processes and components in both the self-driving stack and the underlying truck platform. Should any of these critical components fall outside the acceptable performance parameters, the Kodiak Driver automatically executes a fallback plan, safely pulling the truck over to the side of the road. This fallback capability is critical for promoting safe operations. Our perception system also uses a proprietary, patent-pending system that leverages generative AI-based VLMs to identify and respond to novel, complex edge case scenarios that can be a challenge for more traditional perception techniques.

Kodiak’s Hardware Solution

Guided by its ecosystem-first approach, Kodiak has developed differentiated, modular hardware to power its driverless operations. This hardware includes:

1.SensorPods—Kodiak’s proprietary SensorPods are mirror-mounted self-contained modules that contain cameras, radars and LiDARs. With two SensorPods per vehicle, the Kodiak Driver maintains overlapping fields of view. Kodiak’s patent-pending quick-disconnect attachment is designed to maximize uptime and utilization by enabling fast swaps with minimal training.

2.Redundant Architecture—Kodiak’s modular driverless hardware platform integrates redundancy into all safety-critical functions, including steering, braking, power and compute. This redundancy ensures the Kodiak Driver can maintain safe and reliable driverless operations, even in the event of a component failure.

3.Physical AI Compute—The Kodiak Driver’s physical AI compute platform includes GPUs optimized for automotive applications and a hardened, military-spec compute designed to withstand the harsh environments where the Kodiak Driver operates. The compute platform also includes automotive processors and in-vehicle networking interfaces developed by NXP Semiconductors, which enable critical functions such as vehicle performance monitoring and on-vehicle power management.

4.Actuation Control Engine (“ACE”)—At the core of our safety architecture is the ACE, a custom-designed computer that manages vehicle actuation independently from the main autonomy system. If any safety-critical component of either the Kodiak Driver or the underlying vehicle platform fails, the ACE steps in to execute a safe fallback maneuver and bring the vehicle to a controlled stop.

Commercialization Strategy

Kodiak’s commercialization strategy has always been focused on meeting our customers’ needs. With a large truck driver shortage, an aging workforce and high turnover rates, our customers have consistently told us that they need a flexible DaaS solution. The DaaS solution is expected to help them supplement their existing workforce, grow their fleet and increase their asset utilization. With customer needs in mind, we first implemented with Atlas in December 2024, and expect to grow, our DaaS revenue model by licensing our solution to long-haul trucking, industrial trucking, and defense customers. We offer different license fee structures to align with our customers’ preferred driver cost approach, such as per-mile or per-vehicle. This flexibility is critical in providing our customers with a solution that works for them.

This flexible and customer-centric business model is matched by the versatility of the Kodiak Driver to meet the operational needs of our customers. The Kodiak Driver’s underlying common technology platform and independence from HD maps is expected to enable commercialization across different use cases, including long-haul trucking, industrial trucking, and off-road defense applications.

The initial commercial launch of our recurring revenue-generating driverless trucking solution in December 2024 focused on the industrial trucking sector. The 24/7 nature of oil and gas wells, the acute driver shortage and safety risks in the Permian Basin, and the remoteness of the environment made the oil and gas logistics market ripe for automation. Next, we plan to commercialize our technology in the long-haul trucking sector. We have built an extensive autonomous freight network of approximately 24,000 miles that runs across the southern United States. Following the expansion of our safety case to cover the long-haul operating domain, we plan to launch our long-haul driverless deployment initially in Texas. We believe Texas is an attractive launch market because it is one of the largest freight markets in the United States, and has a generally favorable business and regulatory environment, along with a relatively moderate climate.

In our long-haul operations, we are currently operating revenue-generating commercial operations with Kodiak-owned autonomous trucks. Since 2019, we have hauled commercial freight for our extensive customer base and, we currently offer regular service between our Dallas hub and Houston, Oklahoma City and Atlanta and recently expanded our service offering to El Paso. We have also previously operated consistent service between Dallas and San Antonio, and piloted deliveries between California and Florida.

We partner with commercial upfitters to integrate the Kodiak Driver hardware into commercially available trucks. This includes our partnership with Roush Industries, Inc. (“Roush”), a leading product development supplier serving the mobility, aerospace and defense industries. We believe this upfit-first strategy is the best approach for the current state of the market, which does not yet offer driverless-ready trucks directly from OEMs at scale. Partnering with Roush also gives us strategic flexibility to make continuous incremental improvements to the Kodiak Driver hardware as new technologies and techniques become available. In the long run, we believe integrating our technology into truck OEM manufacturing processes will help us achieve additional economies of scale, further reduce the cost of our system and enable significant scale of production.

We believe our customer-centric approach is a key differentiator. Rather than offering a one-size-fits-all approach, we closely collaborate with our customers through our Partner Deployment Program (“PDP”). We have refined our PDP over several years to ensure the ease of integration of the Kodiak Driver into our customers’ operations. This structured program begins with: (i) assessing the customer’s freight network to identify optimal deployment opportunities; (ii) designing end-to-end solutions that address customer needs; and (iii) defining operational workflows, including maintenance services and system integrations. This enables us to prove the viability of our technology and our commitment to customer success by hauling customer freight with Kodiak-owned autonomous trucks. As we scale our DaaS model beyond the industrial domain, we expect Kodiak-owned autonomous trucks deployed on behalf of our customers will be replaced by customer-owned trucks powered by the Kodiak Driver.

Competition

We believe our main competitors are other AV technology developers, particularly those who focus on long-haul trucking, industrial trucking and defense.

The principal competitive success factors in AVs for long-haul trucking, industrial trucking and defense, include:

•market reputation;

•safety and reliability;

•technology quality;

•team quality;

•go-to-market approach;

•commercial traction;

•capital efficiency; and

•intellectual property portfolio.

We believe we compare favorably to our competitors on most, if not all, of these categories, due to our demonstrated commercial traction, differentiated technology, strong safety culture and experienced team.

Our Growth Strategy

We plan to continue to build on our position as a leading AV technology provider through further commercialization and scaling of our commercial deployments and the continued development of the Kodiak Driver.

Domestic Expansion

Since our inception, we have focused on building a solution that solves customer pain points and easily integrates into our customers’ existing networks. We launched our driverless solution in the Permian Basin in West Texas and Eastern New Mexico with Atlas in December 2024 and expect our deployment with them to scale in 2026 and beyond. In addition, we plan to launch driverless highway operations by the end of 2026 in the southern United States, following the completion of key systems and safety engineering work required to validate the expansion of our safety case to encompass long-haul operations. Once we complete our long-haul safety case, we will then aim to strategically grow our geographic footprint to freight lanes across the country. We expect to expand nationwide over the course of the decade to meet the demands of our customers and as we validate our technology in additional operating environments.

We also see opportunities in the defense sector, particularly given the Pentagon’s increasing preference for commercial off-the-shelf technologies, rather than home-grown or defense-specific solutions. The Pentagon has stated that it views autonomous technologies as a priority area for investment, particularly as a means of maintaining combat superiority against adversaries that can field greater manpower.

International Expansion

While we are focusing our commercial operations in the United States in the near term, we expect to expand to key international markets that share certain similarities with the United States such as labor shortages, supportive regulatory environments and comparable roadway environments. We expect these jurisdictions may initially include Australia, the Middle East, Canada and Europe, subject to ongoing customer and prospective customer discussions, market analysis and trade policy considerations. Additionally, we see defense sector opportunities with allied militaries in Europe, given instability in the region.

Intellectual Property

Our success and competitive advantage depend in part upon our ability to develop and protect our core technology and intellectual property. We own a portfolio of intellectual property, including patents and patent applications, trademarks, domain registrations, confidential technical information including trade secrets, technical data, curated data sets and expertise in the development of software and hardware for AVs.

We have filed applications and paid certain fees to register and maintain our patents and trademarks, to secure and protect our intellectual rights and defend against third parties who may infringe on our intellectual property rights. We also rely on trade secrets to protect our AI model architecture, design and manufacturing know-how and individual model structures and model parameters that have been developed within our proprietary perception, mapping, planning and control systems with curated data structures. Additionally, we take commercially reasonable steps designed to secure and protect our intellectual property rights, including through agreements with our commercial partners, supply-chain vendors, employees and consultants, and by maintaining our software in an access-controlled proprietary cloud environment.

As of December 31, 2025, we own over 150 patents and pending patent applications, including filings in U.S. and foreign jurisdictions. In addition, we have two registered U.S. trademarks and one registered foreign trademark. Our patents and patent applications cover a broad range of technologies relevant to self-driving vehicles across different operating domains, including long-haul trucking, industrial and defense related domains.

Regulatory Landscape

There is currently no comprehensive federal regulatory framework governing the deployment of driverless trucks. We are able to operate our driverless trucking business today as regulators continue to work towards developing regulatory frameworks specifically for AV deployment.

At the state level, the regulatory landscape continues to evolve. As of December 31, 2025, 24 states have enacted legislation explicitly permitting the deployment of driverless trucks. Critical freight corridors, including the I-10 and I-20

corridors stretching from Arizona to Georgia and Florida, are located in states with supportive legal frameworks in place. This has enabled us to pursue and expand our operations across key commercial routes critical to long-haul trucking. While California currently does not allow for AV truck deployment, in April 2025, the state launched a rulemaking process that we expect to lead to regulations allowing for driverless truck deployment in the near term.

While these state laws differ, they generally allow for driverless deployment, so long as vehicles are able to follow the rules of the road and are able to achieve minimal risk conditions. Differences in requirements, such as first responder interaction protocols and insurance standards, create compliance complexities.

At the federal level, the United States Department of Transportation (“USDOT”), through both the National Highway Traffic Safety Administration (“NHTSA”) and the Federal Motor Carrier Safety Administration (“FMCSA”), continues to study and issue guidance related to autonomous vehicles. Both agencies have devoted considerable resources to understanding how automation will change roadway operations and safety. Secretary of Transportation Sean Duffy has pledged to create a federal framework for AV deployment, and in April 2025, the USDOT released the first installment of that framework. Additionally, in October 2025, USDOT issued a waiver that allows AV truck operators to use flashing warning beacons as a replacement for reflective warning triangles, providing a solution to a core regulatory challenge facing AV trucking.

Current federal AV regulations largely rely on existing authorities. This includes 2018’s AV 3.0 guidance, which specifically states that “Going forward FMCSA regulations will no longer assume that the commercial motor vehicle driver is always a human or that a human is necessarily present onboard a commercial vehicle during its operation.” Further, in July 2021, NHTSA issued Standing General Order 2021-01 (“SGO”), which requires AV companies to report any crash that “results or allegedly results in any property damage, injury, or fatality.” This SGO has been amended twice in the ensuing years, most recently in April 2025. NHTSA both posts this data on a publicly available portal and uses it to identify potential safety defects that may necessitate a recall under the National Traffic and Motor Vehicle Safety Act of 1966 (the “Vehicle Safety Act”).

In the United States, manufacturers must self-certify that their vehicles comply with the FMVSSs, promulgated by NHTSA. Companies such as Kodiak self-certify continued compliance with the FMVSSs, while NHTSA retains authority to oversee safety through its defect and recall powers. Further, we and our customers must comply with FMCSRs regarding the safe operation of commercial vehicles.

We play a central role in educational and advocacy efforts related to AV deployment. We belong to numerous industry associations and trade groups, including the Autonomous Vehicle Industry Association, the ATA, Partners for AV Education, the Consumer Technology Association, the Permian Road Safety Coalition, the Commercial Vehicle Safety Alliance and more. At a state and local level, we belong to over a half dozen state trucking associations and engage closely with state and local regulators to educate them about the opportunity for AV trucking to improve safety and grow the economy. We co-chair the Freight Subcommittee of the Texas Connected and Autonomous Vehicle Task Force and have twice testified before the Texas House Transportation Committee. We will continue to work closely with federal and state policymakers to strengthen an already robust regulatory picture and deploy our technology.

As the regulatory environment continues to advance, our business may need to evolve accordingly. For example, additional state-level requirements or new federal standards could require operational or technical adjustments. We proactively engage with policymakers and regulators to help ensure the regulatory frameworks support safe and scalable autonomous deployment. We are hopeful that, in the near future, USDOT will issue additional regulations that will create further regulatory certainty for the AV industry and the U.S. Congress will pass a federal legislative framework for AVs.

Corporate Social Responsibilities and Sustainability

Safety

We have a deep commitment to making driving safer. Delivering on this commitment is central to how we aim to make a positive contribution to our communities and our country. Safety is the foundation of everything we do at Kodiak. It is a responsibility that we take seriously and has been a key focus of Kodiak since day one. We also firmly believe that safety requires a team approach, and we have worked closely with federal and state regulators, law enforcement and our industry partners to ensure that our vehicles are fundamentally safe. This commitment is also paramount in our defense business, where we are helping to remove service members from harm’s way in some of the most dangerous jobs in the U.S. military.

Sustainability

We believe that autonomous trucks will help improve fuel efficiency, through less idling, more efficient routing, fewer non-revenue producing miles and more efficient driving. We believe these improvements will facilitate reduced customer emissions, while lowering operating costs for our customers.

Our Culture

Kodiak’s team members are the collective force that drive innovation and organizational excellence. Kodiak has established a clear purpose that employees, partners and customers identify with and are passionate about.

This fosters a collaborative environment that we believe delivers personal and professional satisfaction. By focusing on employee engagement, collaboration, accountability and overall fulfillment, Kodiak is able to attract and retain top talent, which we believe creates a high-performance culture. This in turn allows us to leverage the talent and knowledge of our workforce, contributing to our overall momentum and success.

We’re strongly driven by our company values:

•Safety—Safety first and always. Reinforce safe choices as a community. Behave responsibly.

•Integrity—Seek the truth, especially when it’s inconvenient. Speak with passion, but respect diverse points of view to move towards execution.

•Innovation—Focus, efficiency and elegance. Learn from experience. Implement breakthrough solutions to solve meaningful problems.

•Teamwork—Seek out best-in-class collaborators. Embrace the strength of different perspectives-across teammates, customers and the motoring public.

•Trust and Accountability—Deliver smooth experiences wherever possible. Build trust through action. Be consistently and constantly accountable.

Employees

As of December 31, 2025, we had 341 full time employees and one part-time employee. None of our employees are represented by a labor union, and we consider our employee relations to be in good standing. To date, we have not experienced any work stoppages.

Available Information

Our website is located at www.kodiak.ai, and our investor relations website is located at investors.kodiak.ai. Copies of our Annual Report on Form 10-K, Quarterly Reports on Form 10-Q, Current Reports on Form 8-K and amendments to these reports filed or furnished pursuant to Section 13(a) or 15(d) of the Securities Exchange Act of 1934, as amended, are available free of charge on our investor relations website as soon as reasonably practicable after we file such material electronically with or furnish it to the Securities and Exchange Commission (the “SEC”). The SEC also maintains a website that contains our SEC filings at www.sec.gov.

We announce material information to the public about us, our products and services and other matters through a variety of means, including filings with the SEC, press releases, public conference calls, webcasts, and the investor relations section of our website (investors.kodiak.ai) in order to achieve broad, non-exclusionary distribution of information to the public and for complying with our disclosure obligations under Regulation FD. The contents of our websites and corporate reports mentioned herein are not incorporated by reference into this Annual Report on Form 10-K or in any other report or document we file with the SEC, and any references to our websites or the contents of our websites are intended to be inactive textual references only.