Observed Signal · Jun 27, 2026 · Product Launch · Source: techcrunch · Impact: 4/5 · Sentiment: Neutral
Asian Startups Launch Mythos-Like AI Models
Two Asian AI companies — Tokyo-based Sakana AI and Chinese cybersecurity firm 360 — this week launched new large AI models (Sakana’s Fugu and 360’s Tulongfeng and Yitianzhen) positioned as local alternatives to Anthropic’s Mythos/Fable after a U.S. export restriction limited Anthropic’s global access. Sakana presents Fugu as a frontier, agent-capable model tailored for Japanese language and customers, while 360 released vulnerability-discovery and cyber‑defense tools. The moves come amid concerns that U.S. export controls can remove access to leading models overnight, prompting regional providers to emphasize orchestration and local-language optimisation as hedges against concentrated AI supply. Anthropic reported a run‑rate revenue above $47 billion in May 2026, underscoring the commercial scale of the models affected by the export order.
U.S. export controls on powerful AI models are shifting supply dynamics and prompting regional model launches; this policy-driven market change has broad implications for access to foundation models and national AI infrastructure.
Track Anthropic Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Sakana AI (Tokyo) launched Fugu, a frontier AI model designed for agents and local language optimisation.
- Chinese cybersecurity firm 360 unveiled Tulongfeng (vulnerability-discovery) and Yitianzhen (cyber defence/incident response) AI tools.
- The U.S. government issued an order two weeks earlier preventing Anthropic from providing Mythos and Fable models to non-U.S. users.
- Anthropic reported a run-rate revenue exceeding $47 billion in May 2026.
- Sakana was co-founded in 2023 by former Google researchers Ren Ito, Llion Jones and David Ha.
Connected Companies & Entities
6 Entities mapped“On Wednesday, Chinese cybersecurity firm 360 reportedly unveiled Tulongfeng, an AI tool it says can go head-to-head with Anthropic’s Mythos....”
“Earlier the same week Sakana AI, a Tokyo-based AI startup launched Fugu, a model named after the Japanese word for blowfish....”
“A spokesperson at Sakana AI told TechCrunch that release of its new model was “entirely coincidental,” yet that hasn’t stopped it from capit...”
“Sakana, co-founded in 2023 by former Google researchers Ren Ito, Llion Jones and David Ha, makes affordable generative AI models......”
“According to Reuters, 360’s founder Zhou Hongyi described vulnerability-finding AI as a national strategic asset......”
“According to Reuters, 360’s founder Zhou Hongyi described vulnerability-finding AI as a national strategic asset......”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Asian AI Models Tulongfeng and Fugu Challenge Mythos 5
The Wall Street Journal reports that Chinese AI models are closing the performance gap with leading US systems in cybersecurity tasks. Security researchers told the WSJ that Zhipu AI’s GLM-5.2 can match US models at finding software vulnerabilities and in some Semgrep tests even outperformed Anthropic’s Claude Opus 4.8; with additional instructions both Opus 4.8 and GLM-5.2 reached performance comparable to Anthropic’s Mythos. GLM-5.2 is an open-weight model that can be downloaded and run on private hardware. Chinese firms are also releasing specialized tools — 360 Security Technology unveiled the Tulongfeng vulnerability-finding model — and Microsoft is reportedly considering offering Chinese models on its platforms. The article notes US restrictions on access to top Western models (OpenAI’s GPT-5.6 access limits) amid concerns about national security and competitive dynamics.
White House AI Crackdown Helps Chinese Models Close Gap
The White House's export-control actions temporarily limited access to leading U.S. AI models, notably constraining Anthropic and prompting OpenAI to slow rollout of new models after a government request. Regulators allowed Anthropic to release Mythos 5 to select companies and agencies following a two-week shutdown, while Fable 5 remains unavailable. Meanwhile, Chinese firms such as Zhipu have released open-weight models (GLM 5.2) that researchers say match U.S. frontier labs on some cybersecurity benchmarks. Cost and efficiency advantages have prompted some startups and large firms to test or adopt Chinese open-weight alternatives, raising commercial interest but also cybersecurity concerns about models that can be run on-premise. The policy move could accelerate competitive shifts in AI development, enterprise sourcing and national-security debates over model and chip exports.
Rising AI Costs Renew Interest in Open-Source Models
Rising inference costs and geopolitical frictions have revived industry interest in open-source AI models as a lower-cost, more controllable alternative to frontier proprietary systems. The article reports that Chinese labs have pulled ahead since DeepSeek’s 2025 breakout, with open models such as Qwen and Kimi becoming common foundations for startups. Anthropic’s Mythos family faces export-control scrutiny, adding urgency to alternatives. Poolside — a $12 billion model maker backed by Nvidia, Bain Capital and DST — has publicly refocused on open source and released Laguna XS.2 in April, a model tailored for agentic coding, according to co‑founder and co‑CEO Eiso Kant. Investors including Michael Mignano (USV) and Bill Gurley say startups and companies are rethinking costly “tokenmaxxing” and diversifying model usage to control spend.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.