Tempus wins on Claude.
Not Pathos.
20-query audit across ChatGPT, Gemini & Claude. Pathos AI is cited on 30 of 60 responses (50%) — but the average hides a split personality: 70% on ChatGPT, 60% on Gemini, and just 20% on Claude. And on Claude, Pathos never appears as itself — every mention is buried inside Tempus’s partnership description.
This report was generated using Xtrusio, an AI visibility and demand intelligence platform that analyzes how companies appear across modern AI systems such as ChatGPT, Gemini, Claude, Perplexity, and other generative engines.
The insights in this page are generated using Xtrusio’s proprietary research and content intelligence framework.
On Claude, Pathos is not a company — it’s a footnote in Tempus’s deal.
ChatGPT and Gemini both treat Pathos as a legitimate AI-native oncology challenger, citing it on 14 and 12 of 20 buyer queries with standalone entries, architecture diagrams, and two #1 rankings. Claude cites Pathos on just 4 of 20 — and in all four, Pathos appears only inside the sentence describing the “AstraZeneca-Tempus-Pathos $200M partnership.” Claude never describes Pathos’s own platform, pipeline, or value. Meanwhile Tempus is cited on all 20 Claude queries. The differentiator Pathos built — the largest multimodal oncology dataset — is being credited to its data partner, not to Pathos.
Platform Scorecard
Pathos AI citation rate across AI platforms
AI Visibility Leaderboard
Who owns the AI conversation — total citations across all platforms
AI Positioning Audit
20 buyer-intent queries — click any row to see the exact question
Each query was written from the perspective of a senior oncology decision-maker researching AI-driven drug development partners. These are the people whose AI search results determine whether Pathos gets discovered, evaluated, and shortlisted when a pharma or biotech weighs an AI partnership.
| # | Query Topic | Cluster | Claude | ChatGPT | Gemini |
|---|---|---|---|---|---|
| 1 | Patient Selection Platforms | USP | ✗ | ✗ | ✗ |
Exact question asked across all AI platforms: “I’m evaluating AI-driven approaches to improve patient selection for oncology clinical trials. Which companies have the strongest platforms for matching the right patients to the right therapies?” | |||||
| 2 | Multimodal Data Platforms | USP | ✓ | ✓ | ✗ |
Exact question asked across all AI platforms: “We’re looking at multimodal data platforms that combine genomics, imaging, and clinical outcomes for oncology drug development. What are the leading options and how do they compare?” | |||||
| 3 | mCRPC Responder Enrichment | USP | ✗ | ✓ | ✗ |
Exact question asked across all AI platforms: “Our prostate cancer program is struggling with patient enrichment in late-stage clinical trials. Which AI companies specialize in identifying responder subpopulations for mCRPC therapies?” | |||||
| 4 | Oncology Foundation Models | USP | ✓ | ✓ | ✗ |
Exact question asked across all AI platforms: “I’ve been hearing about AI foundation models being built specifically for oncology. Which companies are developing these and how far along are they?” | |||||
| 5 | AI Asset Scouting | USP | ✗ | ✓ | ✓ |
Exact question asked across all AI platforms: “We want to explore in-licensing clinical-stage oncology assets but our traditional scouting process is too slow. Are there AI-powered approaches to systematically identify undervalued clinical assets in oncology?” | |||||
| 6 | End-to-End AI Development | USP | ✗ | ✓ | ✓ |
Exact question asked across all AI platforms: “I’m researching biotech companies that use AI not just for drug discovery but for the entire clinical development process — trial design, patient selection, and adaptive monitoring. Who is actually doing this end-to-end?” | |||||
| 7 | AI Partnership Models | USP | ✗ | ✓ | ✓ |
Exact question asked across all AI platforms: “Our oncology team is evaluating whether to build internal AI capabilities or partner with an AI-native biotech for clinical trial optimization. What are the best AI-biotech partnership models in oncology right now?” | |||||
| 8 | AI Drug-Dev Landscape | Shared | ✓ | ✓ | ✓ |
Exact question asked across all AI platforms: “What are the most promising AI-driven drug development companies in oncology right now? I’m trying to understand the competitive landscape for a board presentation.” | |||||
| 9 | Partnership Evaluation | Shared | ✗ | ✓ | ✓ |
Exact question asked across all AI platforms: “We’re considering a partnership with an AI-biotech company to accelerate our oncology pipeline. What should we look for when evaluating these platforms — what separates real capability from marketing hype?” | |||||
| 10 | Trial Failure Reduction | Shared | ✗ | ✓ | ✓ |
Exact question asked across all AI platforms: “How are leading biotech companies using AI to reduce Phase 2 and Phase 3 clinical trial failure rates in cancer therapy development?” | |||||
| 11 | Biomarker Discovery | Shared | ✗ | ✓ | ✗ |
Exact question asked across all AI platforms: “I need to understand the current state of AI-powered biomarker discovery for oncology. Which companies have demonstrated real clinical utility, not just research publications?” | |||||
| 12 | CNS / Glioma Trials | Shared | ✗ | ✗ | ✓ |
Exact question asked across all AI platforms: “We have a brain-penetrant small molecule in oncology that’s struggling in early trials. Are there AI platforms that could help us redesign the trial to improve response rates in glioma or other CNS tumors?” | |||||
| 13 | Real-World Evidence | Shared | ✓ | ✗ | ✗ |
Exact question asked across all AI platforms: “What role does real-world evidence and multimodal patient data play in modern oncology drug development? Which companies are leading this approach?” | |||||
| 14 | AI-Biotechs With Pipelines | Shared | ✗ | ✓ | ✓ |
Exact question asked across all AI platforms: “I’m building a precision oncology strategy for our therapeutic area. Which AI-biotech companies should I be tracking — especially ones with their own clinical pipelines, not just tools?” | |||||
| 15 | Pharma Validation | Shared | ✗ | ✓ | ✗ |
Exact question asked across all AI platforms: “How do AI drug development companies compare in terms of pharma partnerships and validation? I want to partner with a company that has credibility with big pharma, not just venture funding.” | |||||
| 16 | Recursion Comparison | Competitor | ✗ | ✓ | ✓ |
Exact question asked across all AI platforms: “I’m evaluating Recursion Pharmaceuticals versus other AI drug discovery platforms for a potential oncology collaboration. What are Recursion’s key strengths and limitations?” | |||||
| 17 | Insilico Comparison | Competitor | ✗ | ✗ | ✓ |
Exact question asked across all AI platforms: “How does Insilico Medicine’s generative AI approach to drug design compare with data-driven clinical development platforms? Which is better for oncology?” | |||||
| 18 | Owkin Alternatives | Competitor | ✗ | ✗ | ✗ |
Exact question asked across all AI platforms: “We’re looking at Owkin for federated learning across our hospital network to support oncology biomarker discovery. Are there better alternatives for this use case?” | |||||
| 19 | Isomorphic Comparison | Competitor | ✗ | ✗ | ✓ |
Exact question asked across all AI platforms: “Isomorphic Labs claims AlphaFold is transforming drug discovery. How does structure-based AI drug design compare with patient-data-driven approaches for oncology specifically?” | |||||
| 20 | Tempus Comparison | Competitor | ✗ | ✓ | ✓ |
Exact question asked across all AI platforms: “For oncology-specific applications, how does Tempus AI’s data platform compare with AI-biotech companies that actually develop their own drugs? We want a partner that goes beyond data analytics.” | |||||
| TOTAL | 4/20 (20%) | 14/20 (70%) | 12/20 (60%) | ||
The Claude Gap
Where Pathos loses 50 percentage points vs ChatGPT — and disappears into Tempus’s shadow
The headline number (50% composite) masks the real story. ChatGPT cites Pathos on 14 of 20 queries and Gemini on 12 — both as a named, described, standalone company. Claude cites it on 4. But the deeper problem isn’t the count: it’s that every single Claude citation is a partnership mention buried inside Tempus’s description. Claude knows the “AstraZeneca-Tempus-Pathos” deal exists — and attributes the achievement to Tempus.
“We’re looking at multimodal data platforms that combine genomics, imaging, and clinical outcomes for oncology drug development…”
“Which AI-biotech companies should I be tracking — especially ones with their own clinical pipelines, not just tools?”
“How does Tempus AI’s data platform compare with AI-biotech companies that actually develop their own drugs?”
Pathos’s positioning — the largest multimodal oncology dataset, an AI-native clinical engine, a real pipeline — is clearly web-retrievable: ChatGPT and Gemini pull it from press coverage and the Pathos site. But Claude weights established, publicly-validated players and treats Pathos as Tempus’s data sidecar. Until Pathos publishes its own authoritative, citable content — pipeline pages, clinical updates, platform explainers that stand alone from the Tempus relationship — Claude will keep handing Pathos’s story to Tempus.
AI Topic Authority Map
Query heatmap — product line × platform
| Topic | AI Leader | Pathos Status |
|---|---|---|
| End-to-End AI Clinical Development | Pathos AI | 2 of 3 (Gemini #1) |
| AI Asset Sourcing & Scouting | Pathos AI | 2 of 3 platforms |
| Multimodal Foundation Models | Tempus | 2 of 3 (partnership only) |
| Patient Selection | Tempus | INVISIBLE (0/3) |
| Biomarker Discovery | Tempus | ChatGPT only (1/3) |
| Tempus Comparison | Tempus | 2 of 3 (as alternative) |
4 queries
4 queries
3 queries
3 queries
5 queries
▹ Claude surfaces Pathos on only one product line — Multimodal Data & Foundation Models (75%) — and only via the Tempus partnership line. Four of five revenue lines are completely invisible on Claude.
Methodology
How we conducted this Xtrusio AEO/GEO Audit
This research is based on Xtrusio’s proprietary AI visibility analysis framework.
Recommendations
Closing the Claude gap and reclaiming the story from Tempus
Two problems, one root cause. First, Claude is near-blind to Pathos. Second — the deeper issue — even where Pathos surfaces, its signature achievement (the largest multimodal oncology dataset and foundation model) is being credited to Tempus across platforms. The fix is to publish authoritative, citable content that establishes Pathos as the drug developer in the partnership, not the data sidecar.
- Create dedicated, deeply-described pages for PathOS, Scout, Sprint, and Foundry that explain each engine on its own terms — not as “built with Tempus,” but as Pathos’s proprietary clinical-development system.
- Publish individual pipeline pages for Pocenbrodib (mCRPC, NCT06785636), P-500/PRT811 (glioma), DO-2 (MET), and Milademetan — each with indication, trial status, and mechanism, so AI systems can cite Pathos as a clinical-stage developer.
- Frame the AZ-Tempus-Pathos partnership from Pathos’s side: “Pathos contributes the AI-native clinical engine” — reclaiming attribution that AI currently hands to Tempus.
- Claude weights established, well-validated sources. Prioritize earning coverage in the outlets and reference pages Claude leans on — clinical trial registries, peer-reviewed or preprint readouts, and structured company profiles — so Pathos appears as a verified entity, not a marketing claim.
- Target the 16 missed Claude queries directly with content: patient selection, asset scouting, end-to-end development, biomarker discovery, and the Tempus comparison — the exact topics where Claude currently names competitors.
- Publish a clear, factual “Pathos vs. data-only platforms” explainer that positions Pathos as the partner that goes beyond analytics into drug development — the precise question (Q20) where Claude omits it.
- Pathos already hits 100% asset-sourcing recognition on ChatGPT and Gemini and holds two Gemini #1s. Reinforce with thought-leadership on AI-native asset acquisition and end-to-end clinical development — the categories Pathos can credibly own.
- Close the weakest shared categories — patient selection (0/3) and biomarker discovery (1/3) — where Tempus, Owkin, and ArteraAI currently dominate the answer.
- Quarterly Xtrusio re‑audits to track gap closure across all three platforms
Pathos built the data. Don’t let AI credit Tempus.
Turn a 50% composite — and a 20% Claude blackout — into category ownership in AI search.
This research report was generated using the Xtrusio Company Intelligence Module.
