A Los Angeles-based quantum computing startup claimed to achieve a historic breakthrough in fault-tolerant quantum computing, but independent analysis of the company's own published paper reveals the claims don't hold up to technical scrutiny. AIX Global Innovations announced on June 15, 2026, that its software called Seed IQ had achieved fault-tolerant quantum computing (FTQC) on IBM's standard quantum processors, a milestone that would represent one of the most significant advances in the field. However, a detailed technical review shows the company's press release misrepresented what its own research actually demonstrated.

What Did AIX Global Actually Claim?

AIX Global's announcement came with impressive-sounding metrics. The company said it had achieved fault-tolerant quantum computing with "approximately one physical qubit per logical qubit," meaning it could perform quantum error correction without needing massive hardware overhead. The press release also mentioned demonstrations at "distance-3 and distance-5 surface-code quantum error correction," technical terms that suggest robust error-correcting capabilities.

The company reported running 22,500 circuits across eight weeks on five IBM Heron processors, with what it described as "zero detected logical errors." It also claimed twenty-two chemistry simulations all landed within chemical accuracy, suggesting the system could solve real-world problems. CEO Denise Holt declared that "the question is no longer how many more qubits are needed before FTQC becomes possible. Seed IQ makes it possible today through governed execution rather than massive hardware scale".

Denise Holt

Where Does the Technical Problem Lie?

The critical issue emerges when examining what the company's own 103-page paper actually says about how the system works. The paper reveals that while AIX did conduct separate quantum error correction demonstrations at distance-3 and distance-5, those were standalone validation experiments. All the actual computation work, including the chemistry simulations and the "FTQC primitive" demonstrations, ran on what the paper calls a "150-qubit governed encoded register held at substrate distance d=1".

Distance is a fundamental concept in quantum error correction. A quantum error-correcting code of distance d can correct a specific number of errors based on a mathematical formula. At distance-1, that formula yields zero. A distance-1 code cannot correct any errors at all. This isn't a limitation of one particular error-correction approach; it's a mathematical law that applies across all quantum error correction. Without redundant qubits carrying backup information, there is nothing to detect or correct.

AIX acknowledges this limitation in its paper but frames it as a feature, claiming that "Seed IQ governance acts as an operational substitute for code distance." This statement suggests that software can replace the physical redundancy that makes quantum error correction mathematically possible. According to independent analysis, this is equivalent to claiming that software can recover a deleted file from a hard drive that was never backed up; the backup copies are the mechanism by which errors are detected and corrected, and without them, correction is impossible.

How Did the Announcement Mislead the Public?

The press release's presentation of the distance-3 and distance-5 results alongside the claim of a 1:1 qubit ratio created a misleading impression. These results came from entirely different parts of the research campaign. The distance-3 demonstrations used 13 physical qubits, while the distance-5 work used 41 physical qubits. The 1:1 ratio applies only to the distance-1 computation register, where no error correction is actually occurring.

The announcement also used language about "zero detected logical errors" across 22,500 circuits, which sounds like a demonstration of fault-tolerant quantum computing. However, the standard way researchers demonstrate fault tolerance is by showing that logical error rates decrease as code distance increases, a metric called the "below-threshold criterion." Google demonstrated this with its Willow quantum chip, and Quantinuum showed it with trapped-ion systems. AIX did not demonstrate below-threshold scaling for the distance-1 register on which all its major claims depend.

What Actually Happened With Those 22,500 Circuits?

Instead of demonstrating fault tolerance through error-rate scaling, AIX used a multi-stage filtering process. The company's approach included several filtering steps designed to select the best possible outcomes:

• Calibration-aware synthesis: The system selected the best-performing qubits based on current hardware calibration data, a standard technique available in IBM's Qiskit software framework.
• Runtime admissibility checks: Circuits were filtered during execution to remove those that didn't meet certain criteria.
• Three-stage admissibility certification: Results were filtered again at compilation time, execution time, and commit time to remove unfavorable outcomes.
• Heralded post-selection: Measurements underwent additional filtering based on parity stabilizer results, which discards shots that don't meet specific conditions.

This filtering approach is fundamentally different from fault tolerance. Fault tolerance means the system can correct errors that occur during computation. Post-selection means discarding results that don't look good, which doesn't demonstrate the ability to correct errors in real quantum computation.

Who Is Behind This Claim, and Why Does It Matter?

AIX Global's founder and CEO Denise Holt has a professional background centered on technology evangelism and content creation rather than quantum physics research. From 2022 through 2025, she positioned herself as a leading educator on Active Inference AI, producing blog posts, podcasts, and paid certification programs. After the company she was associated with experienced difficulties, she founded AIX Global Innovations. Neither Holt nor co-founder Denis Ovseyenko has published peer-reviewed research in quantum computing, quantum error correction, or quantum information theory. Their prior quantum work consisted of simulations published on their own blog in January 2026, which Holt herself described as "preliminary and simulated, and real hardware validation remains the goal." Five months later, the company announced what it characterized as the most significant result in quantum computing history.

The announcement received coverage from Quantum Zeitgeist, Quantum Computing Report, Yahoo Finance, Morningstar, and AP News via BusinessWire's distribution network. However, as of June 20, 2026, the claim had not been covered by major scientific publications like Nature, Science, IEEE Spectrum, MIT Technology Review, Physics Today, or Ars Technica. No public technical response came from IBM, whose hardware was used for the experiments, and no quantum computing research group has publicly acknowledged or commented on the claims.

What Should Quantum Computing Investors and Observers Know?

This situation highlights the importance of independent technical review in quantum computing announcements. The field has seen genuine breakthroughs from established research institutions and companies with deep quantum expertise. When evaluating quantum computing claims, several factors matter: whether the research has been submitted to peer-reviewed journals, whether independent experts have commented on the work, whether the researchers have prior publications in quantum information science, and whether the technical details in the paper actually support the press release claims.

The quantum computing field is moving rapidly, with real progress being made by organizations like Google, IBM, Quantinuum, and academic institutions. However, the gap between genuine breakthroughs and overstated claims can be significant. Technical scrutiny, peer review, and independent expert commentary remain essential tools for separating real advances from misleading announcements in this emerging field.

" }