Palantir Aug 20

Rating Company Date 19 August 2026 Buy Palantir North America Reuters Bloomberg Rating Buy PLTR.OQ PLTR US Price target (USD) 200.00 United States Price at 18 Aug -week range 207.18 – 107.27 TMT Software Valuation & Risks

Sovereignty Bootcamp Takeaways Brad Zelnick We had the pleasure of attending keynotes and interactive sessions at Palantir’s Research Analyst second Sovereignty Bootcamp in New York. We came away with a better appreciation for how Palantir’s strategy is meeting the moment, as customers Bhavin Shah, CFA confront rising inference costs, limit reliance on a single frontier model provider, Research Analyst and address concerns around governance of proprietary data. Palantir defines sovereignty as customers retaining control of the broader decision system, Chris Fountain including their data, business logic, permissions, workflows, and ultimately Research Associate outcomes, while remaining free to change the underlying models or infrastructure. The Ontology connects these elements in a unified control layer, enabling institutional knowledge to remain protected and have value compound Daniel Knauff Research Associate within the customer’s environment. Importantly, we heard from CEO Alex Karp directly, who kicked off the event by underscoring the importance of having the alpha generated from an organization’s proprietary data and expertise accrue to Yash Kejriwal that organization, rather than an outside model provider. He argued that Research Associate sovereign AI can help protect and distribute technological and economic power

across institutions more broadly, rather than allowing it to become concentrated among a relatively small group in Silicon Valley.

Customer presentations brought this to life. We saw customer demos spanning vehicle inspection, construction operations, supply-chain management, critical infrastructure planning, and defense deployments at the edge. In one example, an open-weight model operated at “1/70th the cost and six times the speed” of frontier models, with the customer continuing to fine-tune it toward its quality target. Other demonstrations showed AI identifying potential supply-chain delays, accelerating responses to construction-site information requests, and prioritizing infrastructure investments based on operational risk. The common thread was digitizing workflows and capturing judgment, actions, and outcomes so institutional knowledge remains protected and compounds over time.

AIP Evolve is new, and the principal product for operationalizing this vision, dynamically optimizing AI workflows across quality, cost, latency, and control. Evolve can test different models, including frontier, open-weight (NVIDIA Nemotron highlighted multiple times), and proprietary, against a customer’s hierarchical needs. It then identifies opportunities to reroute workloads, rewrite prompts, or replace unnecessary model calls with deterministic code. This provides "model liquidity", or the ability to leverage multi-model strategies without requiring customers to rebuild the surrounding workflow amidst ongoing technological advancements, or pricing and provider policy changes. Palantir’s AutoTune capability within Evolve fine-tunes models using the governed

Deutsche Bank Securities Inc. IMPORTANT RESEARCH DISCLOSURES AND ANALYST CERTIFICATIONS LOCATED IN APPENDIX 1. Deutsche Bank does and seeks to do business with companies covered in its research reports. Thus, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision.

feedback generated by the workflow, helping customers retain proprietary knowledge within their own environment and reduce the need to repeatedly expose sensitive context. In our view, multi-model support, evaluation-driven routing, and controlled fine-tuning address several of the most pressing barriers to scaled enterprise AI.

All in, we reiterate our Buy rating and $200 TP, as Palantir continues to stand apart in its ability to convert AI demand into measurable customer value while delivering a rare combination of growth and…