Legendary AI breakthroughs meet tomorrow’s AI solutions
For six decades, SRI has been hard at work to make AI more efficient, effective, and responsible.
AI innovations for today's hardest problems
We’ve been working on AI since the very beginning. SRI's Artificial Intelligence Center, which launched in 1966 as one of the first research groups devoted to AI innovation, turns 60 this year. SRI invented Shakey (the first robot to navigate the world with AI), delivered critical AI-defined work in areas like natural language processing and collaborative autonomy, and launched (through Siri) the era of virtual personal assistants.
Today, as AI reshapes everything from drug discovery to financial systems to manufacturing and defense, SRI researchers are collaborating to build new AI systems that sense, model, and learn to extend human capabilities and accelerate innovation.
PARC Forum: Beyond LLMs — How AI is reshaping scientific discovery
Join us Oct 1 in Palo Alto or online: A discussion where WEF’s Stephan Mergenthaler and SRI’s David Parekh explore how AI is transforming the scientific discovery process itself—not just speeding it up—across fields like drug discovery and materials science.
Human-centered approach. A focus on system transparency, explainability, reliability, and predictability. SRI is dedicated to advancing AI that makes the world safer, healthier, and more sustainable.
Our AI research and solutions
SECURITY
SRI’s research brings customers a deep understanding of problems where AI meets cybersecurity, privacy, and system design. We're building new solutions for an AI-native information space defined by both new threats and new AI-driven opportunities to enhance security and resilience.
AUTONOMY + COLLABORATION
We design collaboration models for human-robot and robot-robot teams; advance new autonomous sensing, reasoning, and navigation capabilities; and build systems that enable humans to work more effectively with complex autonomous systems and organize humans and machines into complex, collaborative, high-functioning units.
EDUCATION
Helping educators and administrators accelerate learning with AI and approach it responsibly, SRI is providing educational systems with rigorous research, innovative AI-native platforms, and evidence-based technical assistance to support growth and achievement for all learners.
HEALTH
SRI is applying AI capabilities to biomedical research, clinician task guidance, genomics, mental health monitoring, and much more. This work creates treatments and platforms that are changing the world for the better.
INFRASTRUCTURE
From materials selection to design to manufacturing and maintenance, AI is redefining how tomorrow’s physical assets will be imagined, built, and sustained. Numerous SRI projects are helping commercial and government customers seize the future of infrastructure.
COMMUNICATION
SRI’s Speech Technology and Research (STAR) Laboratory brings together engineers, computer scientists, and linguists to build systems for signal processing, data indexing and mining, and computer-aided learning.
PERFORMANCE
AI that accelerates the creation of efficient, low-power, edge-based computing solutions. Including architecting data, software, and hardware into coherent AI systems that deliver both speed and value.
RESPONSIBLE AI
Adressing the critical issues of predictability and reliability of AI-based systems. We're framing core responsible AI questions around whether systems are fair, secure, and accurate.
Case studies
We aren't just developing AI. We're utilizing it in the real-world, with real impact.
CHALLENGE
Rural healthcare options around the country are becoming more limited as hospitals and other care facilities close and consolidate. To respond, ARPA-H’s PARADIGM program is developing a mobile platform that can deliver advanced hospital-level care to every rural county in America.
SOLUTION
SRI is building an intelligent task guidance system for the platform that integrates multimodal AI — including real-time computer vision, speech understanding, and procedural reasoning — to guide, monitor, and adapt to user actions.
IMPACT
Initial testing on a blood draw procedure showed that the SRI-developed system generated impressive user satisfaction (95%) while also greatly increasing procedure accuracy. SRI’s task guidance system is designed to enable any qualified healthcare worker to deliver specialist-level care, anywhere, anytime, through intelligent, real-time guidance.
CHALLENGE
Individualized Education Programs (IEPs) could be a critical source of information about how special education services are intended to operate on the ground. But because they contain personally identifiable information, they are closely protected and have been challenging to leverage for system improvement.
SOLUTION
SRI built an AI-powered platform called SEAMLESS (Special Education Applications of Machine Learning to Enhance Student Success) that allows users to redact sensitive and personally identifiable information in bulk from IEPs and transform their contents into analyzable data.
IMPACT
SEAMLESS is being put to use on behalf of school districts around the country to analyze special education service delivery and identify new opportunities for improvement.
CHALLENGE
Industrial facilities often rely on manual checks and fragmented monitoring across large, hazardous, or hard-to-reach environments. This can slow operations, limit visibility, and increase the risk of missed safety or maintenance issues.
SOLUTION
SRI’s Industrial Vision technology combines high-resolution imaging with edge-based artificial intelligence (AI) to turn visual monitoring into actionable operational insight. The system helps facilities automate tasks such as gauge reading, valve position monitoring, anomaly detection, and technician guidance in real time.
IMPACT
Industrial Vision can improve safety, reduce time spent in high-risk areas, and increase confidence in critical facility operations. By bringing more accurate, continuous monitoring to complex environments, it enables faster decisions and more resilient operations.
CHALLENGE
Speech and media processing in defense, intelligence, and law enforcement often requires accommodating degraded speech signals from highly variable and uncontrolled environments.
SOLUTION
Leveraging decades of research in AI, speech, and natural language understanding, SRI built the Open Language Interface for Voice Exploitation (OLIVE) platform, combining numerous mission-critical speech processing capabilities with image/video analysis, face biometrics, and synthetic/manipulated media detection capabilities.
IMPACT
OLIVE now provides customers in defense, intelligence, and federal law enforcement with real-time and post-mission speech processing for multiple mission types. It operates across languages of high operational interest, often with little available resource and no commercial support, providing capabilities including robust dialect identification, voice biometrics and speaker tracking, and language-independent keyword detection.
CHALLENGE
AI at the edge is often constrained by limited power, compute, and bandwidth, especially in fast-changing environments such as autonomous systems and remote sensing. Traditional AI architectures can struggle to deliver the responsiveness and efficiency these applications require.
SOLUTION
SRI’s NeuroEdge software framework emulates aspects of the human brain to process information more efficiently at the edge. By combining neuromorphic principles with adaptive AI, NeuroEdge helps systems learn and respond in real time while using far less compute.
IMPACT
NeuroEdge’s features contribute to a process that may be 100 or even 1,000 times more efficient than a standard AI edge system configuration, enabling faster decision-making in environments where power and latency matter most. The result is more capable, resilient AI for mission-critical applications across sectors such as defense, transportation, and intelligent sensing.
CHALLENGE
The models that power generative AI applications often produce results that contain errors, which AI researchers refer to as “hallucinations.”
SOLUTION
SRI researchers demonstrated how, by decomposing responses into simpler sub-claims and validating sub-claims individually, they could greatly improve the accuracy of Large Visual Language Models (LVLMs).
IMPACT
SRI’s approach reduced hallucination rates by up to 32% across the five major LVLMs.
SRI’s Open Language Interface for Voice Exploitation has become a critical tool for both government customers and next-generation AI startups.
In the age of disinformation, SRI researchers are helping develop cutting-edge AI tools that can tell manipulated news from real.
With bad information running rampant, the need to assign authorship is essential. SRI is developing the tools to make it happen.
Enabling companies to automatically understand the intonation of the human voice.
The emerging field of edge computing could rein in AI's energy splurge. The SRI NeuroEdge framework aims to further increase edge computing's efficiency and responsiveness.
Novel AI approach integrates data-driven modeling with theoretical reasoning to construct sophisticated deep learning architectures.
SRI researchers are building an AI-driven system designed to upskill clinicians and transform rural healthcare delivery.
Researchers at SRI are developing tools to help clinicians keep a close eye on depression, PTSD, and other mental health issues.
DASL combines machine learning and human expertise to provide insights for a variety of applications.
The model generated responses that were more helpful and honest, and less harmful, than others.
SRI is discovering how AI can organize humans and machines into complex, collaborative, high-functioning units.
Using advanced AI to analyze billions of leaked passwords, SRI researchers reveal the risky human tendencies that make passwords far too easy to guess.
Drawing inspiration from human learning, researchers are guiding chatbots to go beyond mere memorization of statistical patterns to understanding of context.
With funding from DARPA, researchers are building an AI technology designed to work alongside humans to promote more prosocial online behavior.
SRI Education researchers provide administrators and policymakers with the context they need to make informed decisions about special education services.
Researchers have developed a new framework that reduces generative AI hallucinations by up to 32%.
Synfini to commercialize and scale automated compound testing for rapid drug production.
Leveraging SRI navigation technology, Entropy Robotics aims to create safe, resilient autonomous platforms for defense, livestock management, disaster recovery, and much more.
Machine learning algorithms identify how hemorrhagic fever viruses hijack human cells.
To chart the future of human-machine teaming, SRI's COLLEAGUE project is building an AI-based system designed to act as a true collaborative partner.
SRI scientists created a new AI-based approach to robotics that enables mobile robots to effectively communicate with each other — and with their human operators.
SRI education and AI experts collaborate to help students thrive in a data-driven world.
SRI developed a system to make AI more trustworthy, resilient, and interpretable.
Helping systems talk to each other increases patient safety, lowers costs, reduces errors, and streamlines treatment.
SRI's AI innovations are helping define the future of collaboration, education, security, and communications
We have the expertise, technology, and the pipeline to bring innovation into the market. And we're just getting started.
