QUANTUM COMPUTING
ALICE & BOB REDUCE QUANTUM ERRORS RATE
Quantum errors pile up fast, so they need to be fixed as they happen. That’s the job of decoding: a classical computer uses the syndrome measurements to infer the most likely errors, so their effect can be undone when interpreting the results.
But decoding has to hit demanding performance targets in latency and throughput, or quantum computing won’t be feasible. That takes tight coupling between the quantum computer and the GPU.
That’s why NVQLink matters. It’s NVIDIA’s platform for tightly coupling QPUs and GPUs, and we’ve been testing it on the decoding of our LDPC-cat codes, Alice & Bob’s error-correcting codes for cat qubits, using Quantum Machines’ OPX1000 and Open Acceleration Stack.
This first step is a proof of concept, run on synthetic error data: decoding our LDPC-cat codes on a GPU with NVQLink is feasible.
We also found that batching pays off. Decoding many syndromes together makes latency grow sublinearly compared to decoding them one at a time. That’s a good sign for scaling, since bigger quantum computers will have far more errors to handle.
Next: making it fast enough for real-time decoding, so error correction can keep pace with the quantum computer.
AI:
LABEL4.AI: Rennes Deeptech Takes On Deepfakes
Deepfakes, cloned voices, AI-generated images and videos: telling real content from fake is becoming one of the defining challenges of the AI era. LABEL4.AI, a deeptech startup from Rennes, is building the tools to do it. After launching its products commercially in early 2026, the company is now accelerating, API News reports.
Science first
LABEL4.AI’s technology does two things. It detects content that has been generated or manipulated by AI, and it identifies the traces that AI leaves behind. The company combines forensic analysis with digital watermarking, which marks AI-generated images, video, text and audio at the moment they are created (JDN).
The approach builds on years of research at France’s CNRS and Inria. “We have science in our bones,” says CEO Nicolas Bodin Guittard. He co-founded the company in December 2024 with Anthony Level, Vivien Chappelier and Rony Abecidan, among others. The founding team brings together public researchers and former staff of TF1 and Qwant. In 2025 it raised about €1 million from business angels.
From the lab to the market
LABEL4.AI targets sectors where trust in content is critical: media, banks, insurers and fintechs. For a bank or an insurer, a convincing fake document, voice or video can mean fraud. For a newsroom, it can mean publishing disinformation. The startup’s ambition is European and international.
Its momentum was recognised this autumn. LABEL4.AI is one of 69 winners of i-Lab 2026, the French State’s innovation competition for startups built on public research (Digital FrenchNation). The prize rewards its Croma project, a software agent that brings its different detection technologies together and analyses content in real time. Croma runs directly on the user’s own computer, so sensitive data never has to be sent to an outside server.
LABEL4.AI is now also working with the French Presidency’s Digital Lab, bringing its detection technology to one of the institutions most exposed to deepfakes and disinformation.
