AI Tool Corrects Quantum Noise Without Extra Hardware

LTM's AI-QEM is shipped in IBM's Quantum Functions Catalog, with validation reported on circuits above 100 qubits.

Updated on Nov. 12, 2026 in Quantum Computing

Bold flat-color editorial illustration of a tiered cryogenic structure with cables descending from its top against a dark navy field.
LTM's AI-QEM quantum error mitigation tool is now available in IBM's Quantum Functions Catalog, with validation reported on circuits above 100 qubits. AI Illustration. Upload story photo >

LTM has made its AI-QEM quantum error mitigation capability available in the IBM Quantum Functions Catalog, a catalog of partner-built quantum software services. The capability is available now as a circuit function for users of eligible IBM Quantum access plans, so it is shipped rather than research-stage.

Why it matters

Quantum computers are inherently susceptible to noise and errors that can affect computational outcomes, making error mitigation a practical bottleneck for useful results. Packaging a partner-built correction layer inside a public catalog lowers the barrier for developers to test noisy hardware without building their own tooling.

The capability is validated on IBM Quantum computers across a range of circuit sizes, including experiments involving more than 100 qubits, and it requires no additional QPU time, repeated circuit executions, or extra qubits. The source does not report measured error reduction against unmitigated baselines, so the magnitude of improvement remains unquantified.

The players

LTM

LTM is a Global Systems Integrator with a workforce of 87,000 employees across 40 countries, offering quantum optimization, hybrid quantum machine learning, and quantum-safe security programs.

IBM Quantum

IBM Quantum is IBM's quantum computing division, operating quantum hardware and the Qiskit open-source software stack, and hosting the IBM Quantum Functions Catalog of partner-built services.

The details

AI-QEM is an error mitigation layer, meaning it corrects for noise after a quantum program runs rather than changing the hardware itself. It uses AI models to learn the specific error patterns of a given quantum device and then applies corrections through classical post-processing, the ordinary computing that happens on conventional processors after quantum measurements are taken. Because the correction is performed classically and does not add circuit runs or qubits, users operate it inside Qiskit, IBM's open-source quantum programming framework.

Timeline

  1. • LTM announced the availability of AI-QEM in the IBM Quantum Functions Catalog on November 11, 2026.

The Tech Race

The IBM Quantum Functions Catalog gives partners a distribution channel that competes with stand-alone quantum software vendors seeking developer adoption. LTM's listing positions a systems integrator alongside hardware-adjacent tooling as the field races to make noisy quantum devices usable for real workloads.

Developers with eligible IBM Quantum access plans can apply AI-QEM as a circuit function inside their existing Qiskit workflows today. The source does not state pricing or the access tiers that qualify, so users should confirm eligibility directly with IBM.

The takeaway

The trend to watch is whether partner-built error mitigation becomes a standard layer in quantum catalogs, shifting differentiation from hardware access to correction quality. Readers should watch for independent benchmarks comparing AI-QEM's output to unmitigated runs on the same devices.

Further reading

For background on the hardware noise this tool targets, see the Quantum Computing section.