Get your free personalized podcast brief
We scan new podcasts and send you the top 5 insights daily.
As highlighted by Palantir's CEO, corporations are wary of feeding proprietary data into large AI models. They fear AI companies will train on their data to launch competitive products, as seen with Figma, while also struggling to justify the high token costs and measure tangible business returns.
Related Insights
Alex Karp states enterprises are skeptical of AI ROI and fear that feeding data to frontier models from OpenAI and Anthropic trains these platforms to understand and eventually replicate their core business. This IP risk is a major hurdle for adoption, which Palantir positions itself to solve.
As noted by Chamath Palihapitiya, businesses fear deploying major AI models directly, seeing it as letting the 'fox into the henhouse' where their usage data could train a future competitor. This creates a strategic opening for 'harness-first' companies that offer enterprises control and choice over underlying models.
Beyond data privacy, enterprises are concerned that AI agents powered by frontier models will absorb their institutional knowledge. This creates a risky operational dependence where core business learnings are owned and controlled by an external AI company, not the enterprise itself.
A CIO can survive a standard data breach, but a CIO who gives away proprietary company data to an AI model will be fired. This distinction explains the high level of caution from IT leaders, which is rooted in existential career risk, not just resistance to new technology.
Alex Karp argues that companies using third-party frontier models are inadvertently transferring their "alpha"—proprietary data, workflows, and competitive advantage—to the AI labs. He advocates for "AI sovereignty," where organizations own their compute, data, and models to protect their intellectual property.
The conversation around AI in SaaS is maturing. Founders are moving beyond pure excitement and are now raising critical counterpoints, questioning whether customers want their data touching LLMs and identifying situations where *not* implementing AI is the better strategic choice.
Despite massive enterprise spending on AI that fuels hypergrowth for companies like Anthropic, non-tech companies find it difficult to realize tangible value. This creates a conflict where CFOs question the spend while CIOs warn of disruption if they pause.
HubSpot's customers revolted not just because their data would train AI, but because it might be shared with other users, including competitors. This rapid reversal highlights that for enterprise customers, protecting the competitive advantage embedded in their curated data is a far greater concern than the act of AI model training itself.
Mission-critical industries like finance and drug discovery are hesitant to use major LLMs because they don't want to share proprietary data with a 'big brain for all.' This creates a significant B2B market gap for custom, private AI models that can be tailored to specific tasks and datasets without compromising privacy or security.
Businesses are wary of embedding Large Language Models into core processes because they fear providers could drastically increase prices later, creating dependency lock-in. This caution slows corporate adoption and challenges the narrative of rapid, widespread integration, posing a risk to optimistic growth forecasts.
