Back to News
EnterpriseAI Understanding briefing

AstraZeneca renews $200 million AI oncology partnership with Tempus for five years

AstraZeneca has extended its strategic oncology collaboration with Tempus AI, committing $200 million through 2031 to use multimodal data and AI tools for cancer drug discovery and clinical development.

4 min readRead the linked source
Source-provided image accompanying AstraZeneca renews $200 million AI oncology partnership with Tempus for five years
Source referenceSource recorded
Publisher
oncodaily.com
Source type
Linked source — primary-source status has not been established.
ContextUnderstand this in 60 seconds

Key terms

Machine Learning (ML)
Methods that allow systems to learn patterns from data and improve over time.
Foundation Model
A large pre-trained model that can be adapted to many downstream tasks.
Precision
The proportion of predicted positives that are actually correct.

What happened

AstraZeneca and Tempus AI have renewed and expanded their strategic oncology partnership for an additional five‑year term, committing $200 million in fees payable to Tempus. The agreement, signed on 16 November 2026 and disclosed in an SEC filing on 18 November 2026, will run from 1 January 2027 through 31 December 2031, with a possible extra year if the full financial commitment is not met. The renewed deal covers access to Tempus’s Lens multimodal data platform, licensed oncology datasets, AI‑powered research tools, and scientific analysis services to support AstraZeneca’s drug‑discovery and clinical‑development programs. The partnership, originally launched in 2021, has already produced AI‑enabled clinical workflows such as a biomarker‑testing program for non‑small‑cell lung cancer and a multimodal oncology that demonstrated a C‑index of 0.802 for overall‑survival prediction in EGFR‑mutant NSCLC patients.

The renewed agreement was formalized in an SEC filing on 18 November 2026 and will become effective on 1 January 2027. AstraZeneca will pay $200 million in total, with an annual target of $40 million and a minimum of $32 million per calendar year. The fee structure covers both existing Tempus products—such as the Lens multimodal data platform and licensed oncology datasets—and future services that may be developed during the term.

The partnership builds on a 2021 collaboration that initially focused on applying AI and machine learning to de‑identified clinical and molecular data for drug discovery. Since then, the scope has broadened to include clinical‑development support and AI‑enabled care‑pathway tools. Notable milestones include a 2024 pilot of Tempus Next, an AI‑driven platform that flags NSCLC patients who may need molecular testing, and a 2025 agreement to co‑develop a multimodal oncology with Pathos AI.

In July 2026, Tempus delivered the first version of its oncology to AstraZeneca. A proof‑of‑concept analysis presented at the 2026 ASCO Annual Meeting reported a C‑index of 0.802 for overall‑survival prediction in EGFR‑mutant NSCLC patients treated with osimertinib, indicating strong discriminative ability, though no direct clinical benefit was demonstrated.

The agreement includes data‑governance clauses that prohibit re‑identification of patients, contacting individuals whose records are used, and unauthorized data use. These safeguards are intended to protect patient privacy while allowing large‑scale AI training on real‑world oncology data.

Source details: oncodaily.com

Why it matters

The renewal signals a deepening commitment by a major pharmaceutical company to embed artificial‑intelligence and large‑scale multimodal data analysis into its oncology R&D pipeline. By allocating $200 million over five years, AstraZeneca is betting that AI can accelerate target identification, improve patient stratification, and streamline clinical‑trial design—potentially shortening development timelines and reducing costs for new cancer therapies. The partnership also illustrates how de‑identified real‑world patient data can be leveraged at scale while maintaining strict data‑governance safeguards, a model that could become standard across the industry. However, the agreement does not yet provide evidence that AI‑driven approaches have improved patient outcomes; future impact will need to be measured through validated biomarkers, trial efficiencies, and ultimately clinical results.

The scale of the financial commitment underscores AstraZeneca’s strategic belief that AI can materially improve the efficiency of oncology drug development, a sector where R&D costs are high and success rates low.

By integrating multimodal data—clinical records, molecular profiles, pathology images, and treatment outcomes—Tempus aims to create more holistic models that can uncover novel biomarkers and predict patient responses more accurately than single‑modality approaches.

If successful, the partnership could set a precedent for other pharma‑AI collaborations, encouraging broader adoption of AI‑driven research pipelines and potentially accelerating the delivery of ‑medicine therapies to patients.

The partnership also highlights the growing importance of data‑governance frameworks in AI research, as large, de‑identified datasets become central to model training. Effective governance will be critical to maintaining public trust and complying with regulatory expectations.

Interactive Mechanism

Interactive Mechanism: How It Actually Works

Explore the underlying technology behind this development interactively.

Model Parameter Size:8B Parameters
VRAM Required5.5 GBGPU memory footprint
Target HardwareMacBook / Single GPUDeployment tier
Privacy100% Air-GappedLocal device capability
Core takeaway: Small, quantized models (3B–8B) now run directly inside smartphones and laptops with complete data privacy, while mammoth 400B+ models remain the domain of datacenter clusters.
Interactive Concept Check+10 Points
AI Models Explained Quiz

Which component of an AI application is the machine-learning model itself?

What to watch next

Key indicators to monitor include: (1) published research or regulatory filings that credit Tempus AI models with accelerating target discovery or trial enrollment; (2) measurable improvements in biomarker‑testing rates or turnaround times at sites using Tempus Next; (3) any new drug candidates entering clinical trials that cite AI‑derived insights from the partnership; and (4) how data‑governance provisions are enforced as the volume of de‑identified patient data grows. Observers should also watch for any early termination or renegotiation triggers tied to financial milestones, which could affect the partnership’s longevity.

Publication of peer‑reviewed studies that attribute drug‑target discoveries or trial‑design efficiencies directly to Tempus AI tools.

Regulatory filings (e.g., IND or NDA submissions) that reference AI‑derived insights from the AstraZeneca‑Tempus collaboration.

Expansion of the Tempus Next platform beyond the initial 15 provider sites, including any reported improvements in biomarker‑testing uptake or timeliness.

Updates on the performance and validation of the multimodal oncology , especially any prospective clinical trials that use the model for patient stratification.

Any amendments to the agreement triggered by unmet financial targets or early termination clauses, which could signal challenges in realizing the partnership’s AI objectives.

Related guides & quizzes

Found this useful?