Observed Signal · Nov 6, 2026 · Policy Update · Source: Trending Topics (DACH/CEE Innovation & Tech) · Impact: 4/5 · Sentiment: Negative
Baidu's ERNIE 5.5 Flash Base Bars Europe, UK, South Korea
Baidu has released ERNIE 5.5 Flash Base, a 280-billion-parameter open-weight mixture-of-experts language model, on Hugging Face. The model, trained on 33.5 trillion tokens, outperforms DeepSeek-V4-Flash-Base on many benchmarks. However, its license explicitly excludes the European Union, the United Kingdom, and South Korea, prohibiting use, fine-tuning, and distribution in these regions. This marks a shift from its predecessor, which was released under the permissive Apache 2.0 license. The new license requires prominent 'Powered by ERNIE' attribution for commercial products, a one-time notification for companies with revenue up to $50 million, written approval for higher revenue, and a ban on using the model to improve other AI models. The restrictions are likely a response to the EU AI Act and similar regulations, echoing moves by Meta and Apple. Despite this, competitors like DeepSeek and Alibaba offer more permissive licenses.
Baidu's decision to exclude Europe from its latest open-weight model highlights the growing impact of AI regulation on global technology distribution. This trend, also seen with Meta and Apple, could fragment the AI ecosystem and put European developers at a disadvantage, affecting the competitive landscape of AI in the AdTech industry.
Track Baidu Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your Polaris7 workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Baidu released ERNIE 5.5 Flash Base, a 280-billion-parameter mixture-of-experts model, on Hugging Face.
- The license prohibits use, fine-tuning, and distribution in the EU, UK, and South Korea.
- The license requires 'Powered by ERNIE' attribution, notification for revenue up to $50M, and bans using the model to improve other AI models.
- ERNIE 5.5 Flash Base outperforms DeepSeek-V4-Flash-Base on 24 of 30 benchmarks, scoring 93.3 on MMLU.
- The restriction is likely a response to the EU AI Act and other regulations, similar to Meta's and Apple's actions.
Connected Companies & Entities
8 Entities mapped“Baidu has released ERNIE 5.5 Flash Base, a 280 billion parameter language model, on Hugging Face with a license that excludes the EU, UK, an...”
“DeepSeek-V4-Flash, its closest architectural relative with 284 billion parameters, is released under the MIT license, the most permissive th...”
“Alibaba's Qwen models are released under Apache 2.0.”
“Tencent followed this week with Hy4, a large open model of its own.”
“In 2024, Meta barred companies based in the EU from the multimodal Llama 3.2 models, citing the 'unpredictable' regulatory environment.”
“Moonshot AI released Kimi K3.1 Pro one day before Baidu.”
“Apple brought Apple Intelligence to Europe months late.”
“Mistral remains a European alternative, releasing its open models under Apache 2.0.”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
China's AI Export Controls Rely on Top-10 Ranking
China's Ministry of Commerce and Ministry of Science and Technology have proposed adding 40 technologies, including 23 AI-related ones, to the Catalogue of Technologies Prohibited or Restricted from Export. The draft includes a 'globally leading capability' test for large language models, defined as a top-10 global ranking in public assessments. The specific ranking metric is unspecified, creating ambiguity. The compute threshold matches the rescinded US AI Diffusion Rule. Models like Kimi K3 and Qwen3.8 exceed the parameter threshold, but public release may exempt them. Baidu's ERNIE 5.0 also meets the parameter test without public weights. The final list's impact remains uncertain.
Zhipu's GLM-5.3-Flash (Ox Alpha) Released
GLM-5.3-Flash (listed as “Ox Alpha”) — a 320-billion-parameter, natively multimodal model from Zhipu AI (Z.ai) — appeared on OpenRouter/OpenCode in late August 2026 and rapidly became the most-used model on OpenRouter, processing roughly 23 trillion tokens. Bloomberg confirmed the model identity on August 26, and weights were published to Hugging Face under an MIT license. The model claims major efficiency gains (18B active per token, reduced attention compute and a much smaller KV cache), aggressive cost-per-task advantages versus higher-priced competitors, and early dynamic GGUF quantizations enabling very small local footprints. Zhipu's Hong Kong shares rose after the release. The launch also includes a claim that the stealth traffic was served on ~100,000 non-NVIDIA Chinese AI chips, a development with potential infrastructure and geopolitics implications.
China's AI Surge: Innovations Amid Controversies and Concerns
This week Chinese technology companies unveiled several new AI models spanning robotics, video generation and large language models. Alibaba’s DAMO Academy introduced RynnBrain, a robot-focused model with built-in time-and-space awareness demonstrated on tasks like object identification and multi-step manipulation. ByteDance released Seedance 2.0, a text-and-media-to-video generator praised for controllability and production quality but which suspended a feature that produced a person's voice from a photo after consent concerns. Kuaishou rolled out Kling 3.0, a subscriber-access video model claiming extended duration (up to 15s) and native multilingual audio. Separately, Zhipu AI (Knowledge Atlas Technology) published GLM-5, and MiniMax updated its open-source M2.5 model with enhanced agent tooling. Reporters note these releases position Chinese firms as closer competitors to Western video and robotics models and raise questions about content consent and model claims.
Receive all future signals for Baidu & competitors in your briefing
Never miss material M&A deals, leadership changes, account wins, or product rollouts. Polaris7 monitors 20,000+ primary sources 24/7, delivering verified telemetry directly into your executive briefing.