TrustAtlas

EU AI Act-compliant AI vendors

AI vendors that publicly attest to compliance with the EU AI Act, including high-risk classification documentation where applicable.

The EU AI Act (Regulation (EU) 2024/1689) classifies AI systems by risk level and imposes obligations on providers and deployers. High-risk obligations apply in full from August 2026; general-purpose AI rules from August 2025. The vendors below publicly attest to EU AI Act compliance. For high-risk use cases, request the technical documentation, conformity assessment, and EU AI database registration where applicable.

Vendors with EU AI Act

Anthropic
HQ: United States
AI safety-focused company building the Claude model family. Founded by former OpenAI researchers with a mission to develop reliable, interpr…
Score 11.44 · low
Amazon (AWS)
HQ: United States
Cloud infrastructure leader that develops proprietary Titan models and custom Trainium/Inferentia chips while offering multi-model access th…
Score 12.34 · low
Salesforce
HQ: United States
Enterprise CRM leader that combines proprietary AI models (Einstein, CodeGen, xGen) with OpenAI integration for Einstein GPT, embedding AI a…
Score 12.74 · low
Adobe
HQ: United States
Creative software leader that developed proprietary Firefly image generation models trained exclusively on licensed content, while integrati…
Score 13.74 · low
Cohere
HQ: Canada
Enterprise-focused AI company specializing in natural language processing for business applications. Known for retrieval-augmented generatio…
Score 13.79 · low
IBM
HQ: United States
Enterprise technology company offering the watsonx AI platform with proprietary Granite foundation models. Combines own model development wi…
Score 14.11 · low
Microsoft
HQ: United States
Global technology conglomerate that both develops proprietary AI models (Phi series) and deeply integrates OpenAI models across its Copilot …
Score 14.68 · low
SAP
HQ: Germany
Global enterprise software company integrating AI through its Joule AI assistant across ERP, supply chain, and business applications. Combin…
Score 16.63 · low
OpenAI
HQ: United States
Creator of the GPT model family and ChatGPT, one of the most widely adopted AI platforms globally. Operates as a capped-profit entity under …
Score 18.36 · low
Google DeepMind
HQ: United States
Google's unified AI research lab combining DeepMind and Google Brain, building the Gemini model family integrated across Google products and…
Score 18.85 · low
Oracle
HQ: United States
Enterprise cloud and database company offering OCI AI Services with both proprietary AI capabilities and third-party model hosting. Provides…
Score 19.89 · low
Palo Alto Networks
HQ: United States
Leading cybersecurity company integrating AI across its security platform through Cortex XSIAM (AI-driven security operations), Prisma Cloud…
Score 19.89 · low
Mistral AI
HQ: France
French AI company building both proprietary frontier models and open-weight alternatives. A leading European AI company positioned as a sove…
Score 21.81 · moderate
Workday
HQ: United States
Enterprise HR and finance platform integrating AI features across workforce management, payroll, and financial planning. Uses proprietary ML…
Score 22.45 · moderate
Nvidia
HQ: United States
Dominant AI hardware manufacturer (GPUs, DGX systems) that also operates DGX Cloud inference platform and NIM microservices for model deploy…
Score 22.63 · moderate
Scale AI
HQ: United States
Data infrastructure company providing AI training data labeling, model evaluation, and RLHF services. Operates both human-in-the-loop data a…
Score 23.3 · moderate
Aleph Alpha
HQ: Germany
German AI company building sovereign European foundation models (Luminous series, Pharia) with a focus on data sovereignty, explainability, …
Score 24.29 · moderate
ServiceNow
HQ: United States
Enterprise workflow automation platform integrating AI through Now Assist, which combines ServiceNow's proprietary Now LLM models with OpenA…
Score 24.4 · moderate
PolyAI
HQ: United Kingdom
UK-based voice AI company specializing in enterprise customer service voice assistants. Powers call centers for hospitality, financial servi…
Score 24.72 · moderate
Palantir
HQ: United States
Data analytics and defense technology company offering the Artificial Intelligence Platform (AIP) that integrates LLMs with operational data…
Score 25.09 · moderate
Cloudflare
HQ: United States
Edge network and security company that operates Workers AI, an inference platform running open-weight models on its global network. Hosts mo…
Score 25.89 · moderate
Synthesia
HQ: United Kingdom
AI video generation platform that creates professional videos with AI avatars from text scripts. Enterprise-focused tool for training, onboa…
Score 29.28 · moderate
Kyutai
HQ: France
French non-profit AI research lab founded in 2023, backed by Xavier Niel, Eric Schmidt, and Rodolphe Saade with €300M initial funding. Focus…
Score 29.85 · moderate
Stripe
HQ: United States
Global payments infrastructure company using AI for fraud detection (Radar), revenue optimization, and billing intelligence. Integrates mach…
Score 29.97 · moderate
Twilio
HQ: United States
Communications platform integrating AI for intelligent routing, voice intelligence, and customer engagement through its CustomerAI technolog…
Score 31.65 · moderate
Haystack (deepset)
HQ: Germany
deepset builds Haystack, an open-source Python framework for production LLM applications including RAG, agentic pipelines, and semantic sear…
Score 31.69 · moderate
Meta AI
HQ: United States
Meta's AI research division developing the Llama family of open-weight foundation models. Largest contributor of open-weight frontier models…
Score 32.15 · moderate
Humanloop
HQ: United Kingdom
LLM evaluation, prompt management, and observability platform for enterprise AI teams. Collaborative workspace for product managers and engi…
Score 32.76 · moderate
H Company
HQ: France
French AI company (formerly Holistic AI, H by Holistic) building autonomous agents for enterprise workflow automation. Founded by former Goo…
Score 35.71 · moderate

Buyer checklist

Compliance is necessary, not sufficient. Holding EU AI Act is a meaningful baseline, but no certification covers AI-specific risk end-to-end. Layer this on top of vendor-specific diligence — sub-processor disclosure, training-data policy, model card transparency, dependency-chain mapping.

Related compliance frameworks