Samsung and SK Were Behind Anthropic's 1,440 Trillion Won Valuation

\r\n
[Financial News] Anthropic, the artificial intelligence startup, has secured a massive funding round and surpassed OpenAI in valuation. With global memory chipmakers including Samsung Electronics, SK hynix and Micron joining as strategic investors, the race for AI dominance is increasingly spilling over into a battle for semiconductor supply chain alliances.
Anthropic, the developer of Claude, said on the 28th local time that it raised $6.5 billion in its recent Series H funding round and was valued at $96.5 billion after the investment.
That is more than 2.5 times higher than its February valuation of $38 billion. It also exceeds OpenAI's $85.2 billion valuation as of the end of March. OpenAI, however, is still expected to reach a market value of as much as $100 billion if it goes public.
Anthropic said earlier this month that its annualized revenue run rate had topped $4.7 billion. The company said demand for enterprise AI is expanding, and the number of customers using Claude in the workplace is rising rapidly.
Krishna Rao, Anthropic's chief financial officer, said, "This financing will help us meet the unprecedented demand we are seeing, stay at the frontier of research, and bring Claude into more workplaces."
The investment drew particular attention because Samsung Electronics, SK hynix and U.S. chipmaker Micron joined as "strategic infrastructure partners." That means all three of the world's top memory chipmakers have now taken part in Anthropic's funding.
Anthropic said, "The technologies of these companies play a key role in the global supply of memory, storage and logic chips," adding, "This partnership will greatly help us expand computing capacity in line with customer demand."
Industry observers see the investment as symbolic, given that competition over AI model performance is ultimately becoming a race to secure high-performance memory and data center infrastructure. For Samsung Electronics and SK hynix, the expansion of the generative AI market is also expected to translate directly into higher demand for High Bandwidth Memory (HBM) and next-generation memory, aligning their strategic interests.
Anthropic also announced plans to unveil its next-generation AI model. The company said it will release Claude Mythos, its top-tier model that had been withheld from public release over concerns about cyber abuse, within weeks.
Anthropic explained, "A model with this level of capability requires stronger cybersecurity measures before public release," adding, "We are developing the necessary safeguards quickly."
Earlier this month, Anthropic disclosed that it had developed Claude Mythos, an AI model capable of detecting cybersecurity vulnerabilities at an expert level, but postponed its release over concerns that hackers and others could misuse it.
Along with that, the company also unveiled Opus 4.8, a submodel of Mythos and the highest-tier version among its currently available models.
Anthropic said Opus 4.8 significantly reduced the rate of "misaligned behavior," in which a model agrees with prompts from malicious users, while also improving the honesty and stability of its responses.
It also claimed that the model outperformed OpenAI's GPT-5.5 and Google's Gemini 3.1 Pro in major benchmarks covering coding, agent use and financial analysis. In the "Humanity's Last Exam" benchmark, which measures expert-level reasoning, it scored 49.8%, the company said.
Opus 4.8 is available starting today, and its API pricing is unchanged from the previous version, Opus 4.7.
km@fnnews.com Kim Kyung-min Reporter