Business artificial intelligence will continue relying on traditional architectures for the next few years. That’s the conclusion from Gartner, which states that no enterprise-scale AI workload will run on quantum hardware before 2028 and that GPU-accelerated computing will remain the foundation of all relevant production environments.
The consultancy also issues a warning to companies and CIOs: many solutions marketed today as “Quantum AI” do not actually run artificial intelligence on quantum computers, but instead use hybrid algorithms or quantum-inspired techniques that operate on conventional infrastructure.
The key points of Gartner’s forecast in 30 seconds
- Gartner dismisses the idea that enterprise AI will use quantum computers before 2028.
- There is no peer-reviewed scientific evidence demonstrating a quantum advantage in real AI workloads.
- Most current “Quantum AI” solutions use GPUs and classical algorithms.
- The firm recommends keeping budgets for AI and quantum computing separate.
- Chief Information Officers should continue prioritizing GPUs, GenAI, cloud, and cybersecurity over quantum projects.
While manufacturers like IBM, Google, Microsoft, IonQ, and PsiQuantum continue accelerating quantum hardware development, Gartner believes there is still a long way to go before this technology can compete with current AI infrastructures.
“Quantum AI” doesn’t necessarily mean quantum artificial intelligence
One of the main messages of the report is that the market uses the term Quantum AI too broadly.
According to Gartner, many vendors claiming to offer quantum AI solutions are actually referring to very different technologies.
The consultancy distinguishes four clearly different categories.
The first is classical AI, based on deep learning models, transformers, or reinforcement learning run on CPU, GPU, or TPU.
The second category includes algorithms inspired by quantum computing, which borrow concepts from quantum mechanics but operate entirely on traditional hardware.
Third are hybrid models, where small quantum circuits collaborate with classical HPC or AI systems, typically in research projects.
Only the fourth group constitutes genuine quantum AI, meaning algorithms that actually depend on quantum processors for execution.
No demonstrated advantage over GPUs
For Gartner, the issue is that there is still no peer-reviewed scientific evidence showing that a quantum computer can execute enterprise AI workloads better than GPU-accelerated infrastructure.
The consultancy believes achieving this requires resolving several technological challenges simultaneously:
- Much more stable hardware.
- Effective error correction systems.
- New layers of specialized software.
- Algorithms specifically designed for quantum architectures.
Until these four aspects evolve together, quantum computing will remain a research platform rather than a viable alternative for running enterprise AI models.
Gartner: separate budgets for AI and quantum computing
The firm also delivers a clear message to technology leaders.
According to Gartner, mixing budgets for artificial intelligence and quantum computing can create unrealistic expectations.
The reasoning is simple: both technologies have completely different future horizons.
Generative AI already offers measurable improvements in productivity, automation, and customer service with returns that can appear in less than two years.
Conversely, quantum computing has yet to demonstrate equivalent benefits in real AI applications.
Therefore, Gartner recommends treating quantum projects as R&D initiatives, avoiding diverting resources from AI platforms that already deliver business value.
CIOs continue to prioritize GPUs, cloud, and cybersecurity
Another notable point from the report is that quantum computing still doesn’t rank among the top investment priorities for CIOs.
According to Gartner, the largest technology budgets remain focused on areas such as:
- Generative AI.
- AI agents.
- Cloud infrastructure.
- Cybersecurity.
Organizations funding quantum projects do so, according to the consultancy, with significantly lower investments and without expecting short-term economic returns.
What companies should do
Rather than advising to ignore quantum computing, Gartner recommends a pragmatic strategy.
The consultancy suggests leveraging quantum-inspired algorithms that run on GPUs and can improve optimization, simulation, graph analysis, or certain machine learning workloads.
It also advises that any quantum pilot project include, from the start, clear success criteria and conditions for cancellation if it does not demonstrate advantages over classical solutions.
Finally, it encourages monitoring sector developments with a focus on logical qubits, error correction, and control software, rather than solely on the total number of physical qubits announced by vendors.
For Gartner, these will be the true indicators of when quantum computing can start competing with traditional architectures in artificial intelligence.
Frequently Asked Questions
When does Gartner believe enterprise AI will use quantum computers?
The firm considers that no enterprise-scale AI workload will run on quantum hardware before 2028.
Is there already a demonstrated advantage for quantum AI?
No. Gartner states that there is no peer-reviewed result showing a quantum advantage in AI workloads used in production.
What exactly is “Quantum AI”?
Strictly speaking, it is AI that needs to run on quantum hardware. Gartner notes that many current products use the term to refer to hybrid techniques or quantum-inspired algorithms operating on GPUs.
What does Gartner recommend to companies?
Prioritize AI that already delivers value, keep budgets for AI and quantum separate, and implement quantum pilots with clearly defined objectives and success metrics.
Sources:
- Gartner, Gartner Predicts Enterprise AI Workloads at Scale Will Not Run on Quantum Hardware Through 2028.
