$800 Million Acquisition Offer Rejected — Choice to Preserve Technology Direction
AI Chip Self-Production Challenge — What Preparation Do Government and Market Need?
Korean AI semiconductor startup FuriosaAI rejected Meta's $800 million (approximately 1.08 trillion won) acquisition offer — interpreted not as a simple deal failure but as a choice to preserve technological independence, attracting domestic and international tech industry attention. CEO Baek Joon-ho personally informed internal members of the negotiation conclusion. The primary reason: concern that post-acquisition, the company's technology could be limited to Meta's proprietary services — reading a worry that choosing global capital means moving according to their direction.
Technology foundation: FuriosaAI has been developing a new AI semiconductor chip named "RNGD (Renegade)." This chip functions as a "brain" enabling AI to quickly think and judge (infer) — particularly helping large language models like ChatGPT rapidly process and respond in real time. Reportedly successful in performance evaluation tests sufficient to justify independent survival. FuriosaAI plans to bring this chip to market in the second half of this year, receiving expectations of providing an alternative reducing dependence on foreign chips.
Funding challenge: FuriosaAI is reportedly in talks with domestic investors including the Korea Development Bank for approximately 70 billion won in fundraising. TSMC is said to be reviewing investment. Policy context: the government previously hesitated to support domestic AI semiconductors labeled as "inference-only" (insufficient for training). But Meta's interest and acquisition offer proved domestic technology's potential — the government is now showing renewed interest, announcing plans to invest concentrated budgets through 2027 targeting production of domestic training AI chips by 2028. Collaboration possibility: FuriosaAI's rejection doesn't mean complete severance — industry analysis suggests future potential for technology cooperation, co-development, or chip supply partnerships with Meta. "Furiosa's choice was extremely difficult — rejecting a global conglomerate's offer foregoes enormous capital, but also attempts to preserve the company's philosophy and technology vision. For this choice to succeed: government policy support, market investor interest, and above all, competitive product proof must follow."
