Insurers are exploring quantum computing, and public research now includes an Allstate–IBM underwriting optimization experiment. The evidence supports serious investigation. It does not yet establish a broad commercial advantage across insurance, lower consumer premiums, or coverage for a quantum-related incident. This tracker separates announcements, experiments, and demonstrated business outcomes.
How to read this tracker
This is a selected, source-linked starting point, reviewed September 26, 2026, rather than an exhaustive inventory of every insurer's work. We classify an announcement as evidence that an organization has stated an intention or relationship. A published experiment is a stronger, but still bounded, form of evidence. A production claim requires separate operational support.
The articles linked below are primary materials from the organizations or researchers involved. They may have commercial interests. We summarize their stated findings, identify limitations, and add our own interpretation explicitly. A source link is not an endorsement of a vendor, policy, or investment.
May 6, 2025: Allstate joins Chicago Quantum Exchange
Status: partnership announcement. Allstate's newsroom states that it joined Chicago Quantum Exchange as a corporate partner. It describes interest in potential insurance applications, including risk assessment, fraud detection, and customer service. The announcement supports the conclusion that the insurer was organizing research relationships.
What it does not show: a measured improvement in a production workflow, a premium change, or an insurance product covering quantum events. Our interpretation is that a partnership can build skills and access to collaborators. Its business significance must be assessed through later results.
May–June 2026: underwriting optimization preprint
Status: published research preprint. A Quantum Approach to Stochastic Optimization in Insurance Underwriting was submitted to arXiv on May 2, 2026, and revised on June 16. The authors investigate a combined quantum and conventional approach to an optimization problem involving uncertain values and a permitted tolerance for violating a constraint.
The abstract reports experiments on IBM Heron processors involving up to 150 qubits and performance comparable with conventional optimization schemes. Those results are a bounded research finding. The arXiv record is a preprint record; we do not describe it here as an independently verified production deployment or infer peer review from its presence on arXiv.
June 18, 2026: the researchers explain the insurance connection
Status: company explanation of research. IBM's article describes applying the work to home-insurance portfolios with uncertain, correlated losses. It explains a hybrid workflow and compares it with conventional approaches. It also says the method is not ready for problem scales where exact solvers become too slow to be useful.
Our interpretation is that this provides a concrete example of research progressing beyond general enthusiasm. It still leaves open the cost, scale, reliability, and integration questions that would determine commercial value. The company article should be read alongside the paper, not substituted for it.
A separate track: post-quantum security standards
NIST's 2024 cryptography standards are evidence of available defensive standards, not evidence that quantum underwriting works. Security migration and quantum computational applications should therefore be tracked separately. An insurer can strengthen its protection against future quantum attacks without using a quantum computer to price policies.
The distinction also helps readers evaluate marketing language. Quantum-safe may describe a cryptographic feature; quantum-inspired can describe an approach executed conventionally; quantum computing involves quantum hardware. Ask for an exact description rather than assuming these terms mean the same thing.
What would move a result into a stronger evidence category?
Our research review asks for the following information before interpreting an experiment as meaningful business progress. A credible result may still have limits; the aim is to make those limits inspectable.
- A clearly defined insurance problem and representative test data.
- Strong conventional baselines and a fair comparison of the complete workflow.
- Published methods, reproducible results where feasible, and independent scrutiny.
- Accuracy, reliability, operating cost, and scale appropriate for actual use.
- For deployment claims, documented operational outcomes and an explanation of what changed.
What to watch next
Look for follow-up experiments, public benchmark updates, and evidence about operational use. Treat a new product name, research hire, or partnership as its own event. None establishes that a household's premium will fall or that a specific claim will be covered. The most valuable next development is a testable answer to a precise problem.
- A partnership, a preprint, and a production result represent different evidence levels.
- Allstate–IBM research provides a concrete experiment with stated scale and comparison limits.
- Quantum security migration is separate from quantum-based insurance calculations.
Sources & further reading
- Allstate joins Chicago Quantum Exchange — May 6, 2025Allstate · External source
- A Quantum Approach to Stochastic Optimization in Insurance Underwriting — preprint, revised June 16, 2026Allstate and IBM researchers; arXiv · External source
- Allstate explores quantum computing for insurance portfolios — June 18, 2026IBM · External source
- First three finalized post-quantum standards — August 13, 2024NIST · External source