Organizations
Last reviewed
May 9, 2026
Sources
18 citations
Review status
Source-backed
Revision
v2 · 3,630 words
Improve this article
Add missing citations, update stale details, or suggest a clearer explanation.
Suggest edit
CtrlK
Last reviewed
May 9, 2026
Sources
18 citations
Review status
Source-backed
Revision
v2 · 3,630 words
Add missing citations, update stale details, or suggest a clearer explanation.
See also: Terms, Models, Applications, Companies and Non-profit Organizations
Organizations that build, fund, govern, and study artificial intelligence span private companies, government agencies, university laboratories, non-profit research institutes, and international standards bodies. This page is a top-level index of the most active groups working in AI and robotics as of 2026, grouped by mission and primary activity. The list is not exhaustive: hundreds of additional firms work on narrower segments such as data labeling, AI infrastructure, niche vertical applications, and academic spin-outs. Use the wikilinks below to reach individual articles with full histories, products, and references.
The boundaries of "AI organization" are fuzzy because most modern technology companies use machine learning somewhere in their stack. For the purposes of this index, an organization qualifies when one or more of the following is true:
Generic cloud providers are listed when they operate first-party AI services (for example Amazon Web Services hosting Bedrock, Microsoft hosting Azure OpenAI). Pure venture capital firms, holding companies without operating AI products, and consultancies are excluded from this gateway.
The phrase frontier lab has come to mean a small set of organizations that train general-purpose foundation models at the largest scale and operate the AI products most users encounter. The group has grown from three or four firms in 2022 to roughly a dozen by 2026, driven by Chinese open-weight labs and well-funded European entrants.
| Organization | Founded | Headquarters | Primary focus |
|---|---|---|---|
| OpenAI | 2015 | San Francisco, USA | GPT family, ChatGPT, Sora, o-series reasoning models |
| Anthropic | 2021 | San Francisco, USA | Claude family, Constitutional AI, safety research |
| Google DeepMind | 2010 (DeepMind) / 2023 (merged) | London, UK | Gemini family, AlphaFold, Gemini Robotics |
| Meta AI | 2013 (FAIR) | Menlo Park, USA | Llama family, open-weight research, Reality Labs |
| xAI | 2023 | Palo Alto / Memphis, USA | Grok family, Colossus supercluster |
| Mistral AI | 2023 | Paris, France | Open and commercial European models |
| DeepSeek | 2023 | Hangzhou, China | DeepSeek V3, R1, V4 reasoning models |
| Cohere | 2019 | Toronto, Canada | Command family, enterprise retrieval |
| Inflection AI | 2022 | Palo Alto, USA | Pi assistant, partially absorbed by Microsoft in 2024 |
| Character.AI | 2021 | Menlo Park, USA | Roleplay assistants, partially absorbed by Google in 2024 |
| AI21 Labs | 2017 | Tel Aviv, Israel | Jamba, Jurassic series |
| Reka AI | 2022 | San Francisco, USA | Multimodal Reka Core and Flash |
| Adept AI | 2022 | San Francisco, USA | Agent and computer-use models, key staff hired by Amazon |
| Aleph Alpha | 2019 | Heidelberg, Germany | Sovereign European models for government and enterprise |
| Liquid AI | 2023 | Cambridge, USA | Liquid Foundation Models, MIT spin-out |
| Sakana AI | 2023 | Tokyo, Japan | Evolutionary model merging research |
| 01.AI | 2023 | Beijing, China | Yi series, founded by Kai-Fu Lee |
| Moonshot AI | 2023 | Beijing, China | Kimi assistant, long-context models |
| Zhipu AI | 2019 | Beijing, China | GLM series, Tsinghua spin-out |
| MiniMax | 2021 | Shanghai, China | Hailuo video, ABAB language models |
| Baichuan Intelligence | 2023 | Beijing, China | Baichuan series open and closed models |
| StepFun | 2023 | Shanghai, China | Step-2 multimodal models |
| Black Forest Labs | 2024 | Freiburg, Germany | FLUX image models, founded by Stability ex-staff |
| Stability AI | 2020 | London, UK | Stable Diffusion lineage |
| Cognition AI | 2023 | New York, USA | Devin coding agent |
| Poolside AI | 2023 | Paris / San Francisco | Code-focused foundation models |
| Safe Superintelligence (SSI) | 2024 | Tel Aviv / Palo Alto | Founded by Ilya Sutskever, Daniel Gross, Daniel Levy |
For a deeper view of the safety practices and disclosures these labs publish, see Frontier Model Forum, Responsible Scaling Policy, and the International AI Safety Report.
Most large platform companies run internal AI groups that ship both research and product. These groups frequently overlap with the frontier labs above (for example Meta AI is both a research lab and a division of Meta) but have broader product mandates such as search, advertising, productivity software, or cloud services.
| Organization | Parent company | Headquarters | Primary AI activity |
|---|---|---|---|
| Microsoft | Microsoft Corp. | Redmond, USA | Copilot family, Azure OpenAI, in-house Phi models |
| Alphabet Inc. | Mountain View, USA | Gemini in Search and Workspace, Vertex AI | |
| Apple Inc. | Apple Inc. | Cupertino, USA | Apple Intelligence, on-device foundation models |
| Amazon | Amazon.com | Seattle, USA | AGI org, Olympus / Nova models, Bedrock, Alexa+ |
| Meta | Meta Platforms | Menlo Park, USA | Llama, Meta AI assistant, Ray-Ban smart glasses |
| IBM watsonx | IBM Corp. | Armonk, USA | Granite models, enterprise governance tools |
| Salesforce AI | Salesforce | San Francisco, USA | Agentforce, Einstein Trust Layer |
| Tencent AI | Tencent | Shenzhen, China | Hunyuan models, WeChat AI features |
| Baidu AI | Baidu | Beijing, China | Ernie series, Apollo autonomous driving |
| Alibaba AI | Alibaba Group | Hangzhou, China | Qwen open-weight family, Tongyi assistants |
| ByteDance AI | ByteDance | Beijing, China | Doubao, Coze platform, Seed research lab |
| Huawei AI | Huawei | Shenzhen, China | Pangu models, Ascend chips |
| Samsung AI | Samsung Electronics | Suwon, South Korea | On-device Galaxy AI, Gauss models |
| Naver AI | Naver Corp. | Seongnam, South Korea | HyperCLOVA X, Korean-language models |
| Xiaomi AI | Xiaomi | Beijing, China | MiLM models, smart home and EV assistants |
| iFlytek | iFlytek Co. | Hefei, China | SparkDesk speech and language models |
| Sony | Sony Group | Tokyo, Japan | Sony AI for entertainment, gaming, and imaging |
| Lenovo | Lenovo Group | Beijing, China | AI PC reference designs, hybrid AI services |
The robotics industry has expanded rapidly since 2023 as advances in vision-language models and large-scale reinforcement learning made general-purpose manipulation possible. The category includes humanoid manufacturers, quadruped specialists, surgical and industrial integrators, and autonomous-driving firms.
| Company | Founded | Headquarters | Notable platform |
|---|---|---|---|
| Boston Dynamics | 1992 | Waltham, USA | Atlas, Spot, Stretch |
| Tesla (robotics) | 2021 (Optimus program) | Austin, USA | Tesla Optimus |
| Figure AI | 2022 | Sunnyvale, USA | Figure 02 humanoid |
| Agility Robotics | 2015 | Albany, USA | Digit bipedal worker |
| Apptronik | 2016 | Austin, USA | Apollo humanoid |
| 1X Technologies | 2014 | Moss, Norway | NEO household humanoid |
| Sanctuary AI | 2018 | Vancouver, Canada | Phoenix humanoid |
| Unitree Robotics | 2016 | Hangzhou, China | G1, H1 humanoids and quadrupeds |
| AgiBot | 2023 | Shanghai, China | A1 and Genie humanoids |
| XPENG Robotics | 2020 | Guangzhou, China | IRON humanoid |
| Fourier Intelligence | 2015 | Shanghai, China | GR-1 and GR-2 humanoids |
| MagicLab | 2024 | Shenzhen, China | MagicBot G1 / Z1 |
| Mentee Robotics | 2022 | Tel Aviv, Israel | MenteeBot humanoid |
| Astribot | 2022 | Shenzhen, China | S1 humanoid |
| Noetix Robotics | 2023 | Beijing, China | N2 humanoid |
| Booster Robotics | 2023 | Beijing, China | Booster T1 |
| LimX Dynamics | 2022 | Shenzhen, China | CL-1 humanoid |
| EngineAI | 2023 | Shenzhen, China | SE01, PM01 humanoids |
| Galaxea Dynamics | 2023 | Beijing, China | R1 series |
| Beijing Galaxy General Robot | 2023 | Beijing, China | Galbot G1 |
| Kepler Robotics | 2023 | Shanghai, China | Forerunner K2 |
| Clone Robotics | 2021 | Wroclaw, Poland | Synthetic-muscle humanoid |
| Humanoid (company) | 2024 | London, UK | HMND 01 |
| Hanson Robotics | 2013 | Hong Kong | Sophia social humanoid |
A wider catalog with specifications and release dates lives at Humanoid robot manufacturers.
| Company | Founded | Headquarters | Focus |
|---|---|---|---|
| DEEP Robotics | 2017 | Hangzhou, China | Industrial quadrupeds DR01, DR02 |
| ANYbotics | 2016 | Zurich, Switzerland | ANYmal industrial quadruped |
| DOBOT Robotics | 2015 | Shenzhen, China | Cobot arms |
| Doosan Robotics | 2015 | Suwon, South Korea | Collaborative arms |
| Rainbow Robotics | 2011 | Daejeon, South Korea | RB-Y1 humanoid mobile manipulator |
| PAL Robotics | 2004 | Barcelona, Spain | TIAGo, TALOS |
| SoftBank Robotics | 2012 | Paris / Tokyo | Pepper, NAO |
| Toyota (robotics) | 1933 | Toyota City, Japan | Toyota Research Institute robotics, T-HR3 |
| Bear Robotics | 2017 | Redwood City, USA | Servi service robots |
| Covariant | 2017 | Emeryville, USA | Warehouse manipulation, key staff joined Amazon in 2024 |
| DJI | 2006 | Shenzhen, China | Drones and autonomous aerial systems |
| Hikvision | 2001 | Hangzhou, China | Vision systems and inspection robots |
| Skild AI | 2023 | Pittsburgh, USA | Robot foundation models |
| Physical Intelligence | 2024 | San Francisco, USA | Pi-0 generalist robot models |
| Enchanted Tools | 2022 | Paris, France | Mirokai service robots |
Autonomous-vehicle programs that operate under the AI umbrella include Waymo, Zoox, and Uber (Advanced Technologies Group lineage). LG Electronics (robotics) and DJI build broader hardware ecosystems that intersect AI but are not AI-first.
The AI hardware sector is concentrated around a few foundry partners, dominant accelerator vendors, and a longer tail of specialty silicon startups. TSMC fabricates the leading-edge nodes that almost every accelerator vendor depends on; Samsung Electronics and Intel Foundry compete at trailing nodes.
| Organization | Founded | Headquarters | Notable AI products |
|---|---|---|---|
| Nvidia | 1993 | Santa Clara, USA | H100, B200, GB300 NVL72, Vera Rubin |
| AMD | 1969 | Santa Clara, USA | Instinct MI300X, MI325X, MI355X |
| Intel | 1968 | Santa Clara, USA | Gaudi accelerators, Xeon, foundry services |
| TSMC | 1987 | Hsinchu, Taiwan | 3 nm and 2 nm process nodes used by most accelerators |
| Cerebras Systems | 2015 | Sunnyvale, USA | Wafer-scale WSE-3 engine |
| Groq | 2016 | Mountain View, USA | LPU inference accelerators |
| SambaNova Systems | 2017 | Palo Alto, USA | Reconfigurable Dataflow Units, SambaCloud |
| Cambricon | 2016 | Beijing, China | Siyuan accelerators |
| Apple Silicon | 2020 | Cupertino, USA | M-series with Apple Neural Engine |
| Qualcomm | 1985 | San Diego, USA | Hexagon NPU in Snapdragon, AI 100 server cards |
| AWS Trainium 2 | 2018 (Annapurna) | Austin, USA | Custom training silicon for Amazon Bedrock |
| TPU Ironwood | 2015 (TPU program) | Mountain View, USA | Google's seventh-generation tensor processing unit |
Cloud and infrastructure operators that build, host, or finance the largest training clusters include Amazon Web Services, Microsoft Azure, Google Cloud, Oracle Corporation, CoreWeave, Crusoe, Lambda Labs, Nebius, RunPod, Modal, and Anyscale. Inference-as-a-service vendors include DeepInfra, Fireworks AI, Baseten, Replicate, and Together AI. The Stargate Project is the most prominent multi-organization data-center build-out, announced in January 2025 by OpenAI, Oracle, SoftBank, and MGX.
University AI laboratories produce a large share of foundational research, especially in reinforcement learning, mechanistic interpretability, and robotics. The institutions below are the most-cited.
| Institution | Country | Notable AI groups |
|---|---|---|
| Massachusetts Institute of Technology | USA | CSAIL, Lincoln Laboratory, Schwarzman College of Computing |
| Stanford University | USA | Stanford HAI, Stanford AI Lab, CRFM |
| University of California, Berkeley | USA | BAIR, RAIL, Center for Human-Compatible AI |
| Carnegie Mellon University | USA | Robotics Institute, Language Technologies Institute |
| Harvard University | USA | Kempner Institute, Berkman Klein Center |
| Tsinghua University | China | Institute for AI, Knowledge Engineering Lab |
| Peking University | China | Institute for Artificial Intelligence |
| Zhejiang University | China | College of Computer Science and Technology |
| KAIST | South Korea | AI Graduate School, Robotics Program |
| Waseda University | Japan | Humanoid Robotics Institute |
| Allen Institute for AI | USA | OLMo open models, Aristo science reasoning |
| Microsoft Research | Multinational | Phi family, AI Frontiers |
| Nous Research | USA | Hermes open models |
| Black Forest Labs | Germany | FLUX research and image models |
| The Anthropic Institute | USA | Independent grant-making affiliate of Anthropic |
| Beijing Humanoid Robot Innovation Center | China | Tien Kung humanoid platform |
| Zhejiang Humanoid Robot Innovation Center | China | Wukong robot research |
Industrial research divisions inside operating companies include Microsoft Research, Google Research, Meta AI (FAIR lineage), IBM watsonx, and Apple AI/ML.
AI safety institutes and independent evaluators have grown into a distinct segment since 2023. They focus on dangerous-capability evaluations, alignment research, and policy advice. Most maintain memoranda of understanding with frontier labs that allow pre-deployment testing of new models.
| Organization | Founded | Type | Focus |
|---|---|---|---|
| UK AI Security Institute | 2023 | UK government | Frontier-model evaluation, renamed from AISI in 2025 |
| US AI Safety Institute | 2023 | US government (NIST) | Now Center for AI Standards and Innovation |
| AI Safety Institutes | 2023 onward | Network | Coordinated network including Japan, Singapore, EU |
| METR | 2022 | Non-profit | Autonomy and threat-model evaluations |
| Apollo Research | 2023 | Non-profit | Deception evaluations, scheming research |
| Redwood Research | 2021 | Non-profit | Alignment, control protocols |
| Alignment Research Center | 2021 | Non-profit | Eliciting Latent Knowledge, evaluations |
| Center for AI Safety | 2022 | Non-profit | Public-interest safety research, MMLU, HLE |
| Machine Intelligence Research Institute | 2000 | Non-profit | Foundational alignment theory |
| Future of Humanity Institute | 2005, closed 2024 | Academic (Oxford) | Existential risk, longtermism |
Close-related advocacy and grant-making bodies include Open Philanthropy, 80,000 Hours, Effective Altruism-aligned funders, and LessWrong-adjacent forums. The annual International AI Safety Report chaired by Yoshua Bengio summarizes the state of the field for governments.
The last three years have seen a wave of new government bodies focused specifically on AI. The list below covers the most active offices that publish technical evaluations or set binding rules.
| Body | Country / scope | Mandate |
|---|---|---|
| US AI Safety Institute (CAISI) | USA | Pre-deployment evaluations, technical standards, hosted by NIST |
| UK AI Security Institute | UK | Frontier-model evaluation, AI security research |
| EU AI Office | EU | Implementation of the EU AI Act, GPAI codes of practice |
| Japan AISI | Japan | Coordinated under METI and IPA |
| Singapore AI Verify Foundation | Singapore | AI testing and assurance frameworks |
| Korea AISI | South Korea | Established in 2024 under KISA and KAIST |
| China Academy of Information and Communications Technology (CAICT) | China | Algorithm registry, generative AI rules |
| OECD AI Policy Observatory | Intergovernmental | Policy data, AI Principles |
| G7 Hiroshima AI Process | Intergovernmental | International Code of Conduct for AI Developers |
| United Nations AI Advisory Body | Global | Reports under the Global Digital Compact |
Major multilateral agreements include the Bletchley Declaration (2023), the Seoul Declaration (2024), and the Paris AI Action Summit declaration (2025). State-level US legislation tracked here includes California Senate Bill 53, the Transparency in Frontier Artificial Intelligence Act, signed in September 2025.
Though not exclusive to AI, several standards organizations publish foundational technical specifications that AI products must comply with for procurement, certification, or interoperability.
| Body | Type | Relevant AI work |
|---|---|---|
| ISO / IEC JTC 1 / SC 42 | International standards | ISO/IEC 42001 AI management systems, 23894 risk management |
| NIST | US federal | AI Risk Management Framework (AI RMF 1.0, 2023), GenAI profile (2024) |
| IEEE Standards Association | Professional | P7000 series on ethical system design, P3119 procurement |
| ITU | UN agency | Focus Group on AI for Health, AI for Good |
| W3C | Web standards | Web Machine Learning, ARIA-AT |
| OpenSSF / Linux Foundation | Open source | LF AI & Data, model card and SBOM standards |
| Open Source Initiative | Non-profit | Open Source AI Definition (1.0, October 2024) |
Non-profits and open communities run datasets, model hubs, evaluation suites, and policy programs that the rest of the ecosystem depends on.
| Organization | Founded | Mission |
|---|---|---|
| Hugging Face | 2016 | Model and dataset hub, Transformers library, open-weights advocacy |
| EleutherAI | 2020 | Pythia, GPT-NeoX, evaluation harness |
| LAION | 2021 | Large-scale image and text datasets |
| Allen Institute for AI | 2014 | OLMo, Tulu, Aristo, public benchmarks |
| Linux Foundation | 2000 | Hosts LF AI & Data including PyTorch Foundation |
| Partnership on AI | 2016 | Multistakeholder responsible-AI commitments |
| Frontier Model Forum | 2023 | Industry forum for frontier-model safety, founded by Anthropic, Google, Microsoft, OpenAI |
| Open Philanthropy | 2017 | AI safety and policy grants |
| MLCommons | 2020 | MLPerf benchmarks, AI Safety v0.5 |
Though this gateway excludes pure financial firms, several investment programs are notable for shaping which AI organizations exist. They include Y Combinator's batches of AI-native startups, the Anthropic and OpenAI Startup Funds, Andreessen Horowitz's AI portfolio, Sequoia Capital's seed program, and sovereign-backed funds such as MGX (UAE) and the Saudi Public Investment Fund. Specific named programs and grant initiatives, where they have their own articles, are linked from individual company pages such as OpenAI and Anthropic.
Entries on this page link to dedicated articles whenever one exists; many cells link to articles still in development. To suggest an addition, edit the underlying article rather than this gateway page. Each individual article carries its own references and product details.