Alexander Caedmon KarpCo-Founder & Chief Executive Officer

Karp says the partnership gives customers greater control over AI infrastructure, allowing them to manage their own compute, models, and sensitive data while maintaining security.

He believes trust has become the biggest issue. Companies worry that AI providers may gain access to valuable intellectual property and business knowledge.

According to Karp, organizations should always retain ownership of their data, models, prompts, and AI outputs instead of transferring long-term value to third-party AI providers.

He explains that Ontology connects AI models with enterprise data while enforcing security, governance, and operational accuracy.

Karp argues that AI models alone are not enough. Businesses need an application layer that safely integrates AI into real operational workflows.

Rather than locking customers into one model, Palantir allows organizations to switch between multiple AI models based on performance and security requirements.

Government customers can deploy AI within classified environments while maintaining control over infrastructure and sensitive information.

He argues many companies spend heavily on tokens without creating proportional business value and risk exposing valuable intellectual property.

Organizations should ask who owns the data, where it is stored, whether prompts are retained, and whether AI providers can learn from proprietary information.

Karp believes success comes from combining powerful AI models, secure infrastructure, and business-specific operational software.

He says Palantir focuses on operational deployment, security, and measurable business outcomes rather than simply providing AI models.

He encourages executives to prioritize AI sovereignty, protect intellectual property, and invest in platforms that maximize long-term enterprise value.