← Accueil actu | Salle du marché

Nvidia, Palantir restrict advanced AI model usage over IP and privacy concerns, report claims

2026-09-14 · Source:Tom's Hardware

Nvidia, Palantir restrict advanced AI model usage over IP and privacy concerns, report claims

Nvidia, Palantir and others are limiting advanced AI model usage amid rising 'paranoia' over customer intellectual property, which firms fear could be used to improve Anthropic and OpenAI models.

Nvidia, Palantir and a group of other companies have restricted usage of advanced AI models, according to a report, with privacy concerns cited as the driving consideration. The report characterizes the mood across the industry as rising 'paranoia' over customer intellectual property — a sentiment now shaping how some of technology's most prominent names handle outside AI tools.

At the center of the reported clampdown is a specific worry: firms are concerned that intellectual property in their hands — including material tied to customers — could be leveraged by Anthropic and OpenAI to help improve those labs' own AI models. That fear over customer intellectual property, rather than any technical shortcoming of the systems themselves, is what has reportedly driven Nvidia, Palantir and others to restrict advanced model usage.

The tension at play is a familiar one for enterprises weighing advanced AI: the models offer substantial capability, but engaging with them raises questions about where sensitive material ultimately ends up. For companies whose businesses rest on customer trust and proprietary workflows, the report suggests, the calculus has tilted toward caution, enough that restricting model usage is reportedly now being adopted by multiple major firms.

The significance of the development lies partly in who is involved. Nvidia and Palantir sit among the most prominent companies in the enterprise AI landscape, yet both are reportedly adopting a restrictive posture toward advanced models from Anthropic and OpenAI. When firms this deeply embedded in AI adoption choose to limit usage of frontier systems over privacy and intellectual property concerns, the report's framing of rising 'paranoia' carries added weight.

For AI buyers and investors, the report underscores that intellectual property protection has become an explicit factor in enterprise decisions on advanced model usage. Restrictions at Nvidia, Palantir and other firms signal that concerns about customer intellectual property feeding Anthropic and OpenAI model improvements are now a stated driver of corporate AI policy — the escalating industry 'paranoia' the report describes.

Mots-clés: Nvidia · Palantir · Anthropic · OpenAI · AI Privacy · Intellectual Property


À lire aussi


萬安算交所 AIXX · AIXX · Bourse compute vérifiée · Accéder à la salle