Palantir Technologies, the Denver-based AI software company, delivered a scathing critique of Large Language Models (LLMs) during its Q2 2025 earnings call, while simultaneously reporting record-breaking financial results. The company achieved its first-ever billion-dollar quarter, with commercial revenue in the US nearly doubling to $628 million compared to the same period last year.
Ryan Taylor, Palantir’s Chief Revenue Officer and Chief Legal Officer, didn’t mince words when describing the limitations of LLMs. He characterized them as “at best a jagged intelligence divorced from even basic understanding,” explaining that while LLMs may appear to outperform humans in certain problem-solving tasks, they can make “catastrophic errors no human would ever make” in the next moment. This stark assessment positions Palantir’s approach as fundamentally different from the LLM-centric strategies pursued by many competitors.
Taylor contrasted LLMs with Palantir’s ontology-based approach, which he described as “pure understanding concretized in software.” This methodology uses logic and data to create digital models that replicate how organizations actually function, emphasizing what Taylor called “reality, not rhetoric.” The company’s financial success appears to validate this strategy, with much of the revenue growth attributed to a massive 10-year, $10 billion consolidated contract with the US Army.
Shyam Sankar, Palantir’s Chief Technology Officer, addressed questions about winning the AI race by highlighting favorable regulatory conditions. He noted that the Trump administration’s new AI Action Plan, which promotes AI deregulation, has “taken all the brakes off” and that industry customers are “really excited to get to work.” This regulatory environment appears to be creating opportunities for Palantir’s enterprise-focused AI solutions.
On talent retention, CEO Alex Karp made bold claims about Palantir’s unique position in attracting top AI talent. He emphasized that highly talented individuals who believe in Western values would find unparalleled opportunities at Palantir, stating that after 20 years of interacting with agencies and companies across the West, none offer comparable “time of joining to full agency” like Palantir. Karp concluded with a confident promise: “If you come to Palantir, your career is set.”
Key Quotes
LLMs, on their own, are at best a jagged intelligence divorced from even basic understanding. In one moment, they may appear to outperform humans in some problem-solving task, but in the next, they make catastrophic errors no human would ever make.
Ryan Taylor, Palantir’s Chief Revenue Officer and Chief Legal Officer, delivered this pointed criticism during the Q2 earnings call, positioning Palantir’s ontology-based approach as superior to the LLM-centric strategies dominating the AI industry.
By contrast, our ontology is pure understanding concretized in software. This is reality, not rhetoric.
Taylor contrasted Palantir’s methodology with LLMs, emphasizing that their approach uses logic and data to create digital models of organizational reality rather than relying on probabilistic language models.
If you come to Palantir, your career is set.
CEO Alex Karp made this bold promise when addressing questions about talent retention in the competitive AI market, positioning Palantir as offering unparalleled career opportunities for top AI talent who align with Western values.
The Trump administration’s new AI Action Plan has taken all the brakes off and industry customers are really excited to get to work.
CTO Shyam Sankar highlighted how the administration’s deregulatory approach to AI is creating favorable conditions for Palantir’s enterprise and government-focused AI solutions.
Our Take
Palantir’s aggressive stance against LLMs is both strategically savvy and technically defensible. While the AI hype cycle has centered on generative AI, the company is carving out a differentiated position by emphasizing reliability, structure, and domain expertise over the impressive but unpredictable capabilities of LLMs. Their billion-dollar quarter validates that enterprise customers—especially in defense and government—prioritize mission-critical reliability over conversational fluency.
However, this positioning also reveals potential vulnerabilities. As LLMs improve and companies develop better frameworks for constraining their outputs, the gap Palantir identifies may narrow. The company’s success appears heavily dependent on government contracts and a favorable regulatory environment, which could shift with political changes. Additionally, Karp’s comments about Western values and talent may alienate some potential employees in an increasingly globalized AI talent market. Nonetheless, Palantir has identified a genuine market need that pure LLM approaches struggle to address: turning organizational complexity into actionable software systems.
Why This Matters
Palantir’s critique of LLMs represents a significant counternarrative in an AI industry largely dominated by enthusiasm for generative AI and large language models. As companies like OpenAI, Google, and Anthropic pour billions into developing ever-larger LLMs, Palantir’s success with an alternative approach challenges the prevailing wisdom about the future of enterprise AI.
The company’s billion-dollar quarter and doubled commercial revenue provide concrete evidence that enterprise customers may be seeking more structured, reliable AI solutions rather than the probabilistic outputs of LLMs. This matters particularly for mission-critical applications in defense, government, and enterprise settings where “catastrophic errors” can have serious consequences.
The regulatory environment discussion is equally significant. The Trump administration’s deregulatory stance on AI could accelerate adoption of AI systems in sensitive sectors, potentially giving companies like Palantir—with established government relationships and security clearances—a competitive advantage. This represents a divergence from the more cautious, regulation-focused approach seen in Europe and previously in the US.
For the broader AI industry, Palantir’s positioning suggests that the AI race may not be winner-take-all for LLM developers, and that specialized, domain-specific approaches combining traditional software engineering with AI may capture significant market share in enterprise and government sectors.
Recommended Reading
For those interested in learning more about artificial intelligence, machine learning, and effective AI communication, here are some excellent resources:
Recommended Reading
Related Stories
- OpenAI’s $157B Valuation: Can It Win the Brutal AI Race?
- The AI Hype Cycle: Reality Check and Future Expectations
- Wall Street Asks Big Tech: Will AI Ever Make Money?
- Sam Altman’s Bold AI Predictions: AGI, Jobs, and the Future by 2025
- The Artificial Intelligence Race: Rivalry Bathing the World in Data