Danish Pharmaceutical Giant Novo Nordisk Hacked, Internal AI Training Data Stolen
Danish pharmaceutical giant Novo Nordisk confirmed unauthorized access to its IT systems, with hackers stealing approximately 1.3TB of data, including clinical trial information and proprietary AI models.
The FulcrumSec hacker group claims to have obtained multimodal models, training datasets, source code, and logs related to the AI project "Dragonfly," and has begun leaking samples, attempting to extort $25 million; Novo Nordisk stated that some anonymized data of clinical trial patients was copied, but core operations were unaffected.
Market-wise, investors are concerned about the security of pharmaceutical AI intellectual property, leading to short-term capital outflows from AI-dependent drug development companies like Novo Nordisk, shifting towards safer options and competitors. This incident drives capital towards cybersecurity service providers, putting direct pressure on Novo Nordisk as a major shareholder.
Source: Public Information
ABAB AI Insight
Novo Nordisk previously collaborated with OpenAI and others to accelerate the application of AI in drug discovery and clinical optimization, a path similar to competitors like Eli Lilly who rapidly iterate through external AI partners, but has been accompanied by multiple data security incidents highlighting intellectual property vulnerabilities.
In terms of capital strategy, the company has invested heavily in building AI training datasets to support blockbuster drug pipelines like semaglutide, rather than seeking short-term cash, aiming to maintain a leading position in the GLP-1 field through data advantages, creating a closed-loop resource mobilization from clinical data to molecular design.
Similar cases include early data leaks from tech giants like Meta and theft of clinical data from several pharmaceutical companies. Novo Nordisk is currently in a vulnerable control phase as AI-driven drug development transitions from internal closure to an external ecosystem.
From a structural perspective, this fundamentally relates to regulatory changes; hacker attacks and extortion are forcing the pharmaceutical industry to enhance AI data protection standards, with mechanisms in place for publicly funded research projects to undergo stricter audits, preventing capital from chaotically concentrating in weak data security areas.
ABAB News · Cognitive Law
The more valuable the data, the more fragile the protection: AI training assets have become the new oil; those who cannot protect them provide ammunition to their competitors.
Intellectual property is a double-edged sword: accelerating innovation amplifies the risk of leaks; the era of closure is ending, and transparency and accountability are beginning.
Security costs determine competitive barriers: whoever first turns protection into structure will lock in pricing power for AI drug discovery.