Rajjie Sarmey is Founder and Chief Visionary Officer of FutureProof CXO, advancing governed AI, enterprise strategy and durable value.
getty
Boards and CXOs for several years have been asking two separate questions: How far can our AI go, and when will quantum computing become commercially relevant for us?
But the more consequential question may now sit between them: What happens when AI and quantum computing operate within the same enterprise architecture ecosystem?
Quantum and AI convergence has not yet arrived in full maturity. AI is scaling rapidly, while quantum remains constrained, specialized and workload dependent. For boards and CXOs, fault tolerance, error correction, economics and demonstrable advantage still matter. But waiting for quantum maturity before designing for its convergence risks enterprises architecting for yesterday’s business boundaries.
So where are we today?
AI already performs routine tasks including coding, prediction, optimization and, increasingly, agentic workflows. Quantum is progressing differently. Its potential advantage, however, is new computational approaches for specific categories of problem.
Google Quantum AI’s Willow research demonstrated below-threshold quantum error correction, a critical milestone toward scalable fault-tolerant quantum computing.
IBM is pursuing quantum-classical integration, with IBM and its partners demonstrating quantum advantage in 2026 and targeting a large-scale, fault-tolerant quantum computer by 2029, according to its quantum roadmap.
This progress matters, but it is not broad enterprise-grade production readiness.
The short-term architecture advancement is unlikely to be “quantum replaces classical.” More plausibly, CPUs, GPUs, AI accelerators, HPC and quantum resources will coexist, selected driven by workload economics and suitability.
Where will AI and quantum begin to converge in the future?
The initial convergence is probably bidirectional.
AI can contribute to quantum evolution through calibration, error correction, control mitigation, materials discovery and workflow optimization. One practical example is Zapata Quantum’s collaboration with NVIDIA, which applies agentic AI to automate quantum resource estimation, initially targeting quantum-chemistry applications. Separately, researchers affiliated with Zapata have published a framework for comparing quantum and classical generative models in the search for practical quantum advantage.
These are meaningful R&D progressing signals, but not proof of broader commercial AI-quantum advantage now.
Where might we see the first meaningful advantage leap?
I am specifically monitoring domains where computational complexity already carries enormous economic and/or scientific value for enterprises that are at the forefront, including drug discovery and molecular simulation, materials science, energy, scientific research, logistics and selected financial optimization. Medicine and research are particularly promising: AI-quantum convergence can generate hypotheses and identify patterns, while future quantum resources could help explore molecular and chemical spaces that are difficult to address with classical computing methods.
The target architecture still remains hybrid: AI discovers and orchestrates, classical systems perform computation and quantum is invoked where measurable advantage emerges.
I define this strategic model as “The Architecture of Convergent Intelligence: Intelligence x Agency x Quantum Reach > Consequence.”
AI expands enterprise and businesses accessible intelligence. Agents transform intelligence into action. Quantum may extend the computational frontier for selected problems with economic and value advantage. As these capabilities collide, enterprise consequence and the need for governance can increase.
The board question is not ‘When is quantum arriving?’
Boards should be cautious about technology timelines. The more meaningful questions are about design and architecture.
Can applications consume different compute classes? Is data governed for hybrid workflows? Can we separate economic advantage from technical novelty? Can decisions be engineered across AI models, agents, classical systems and quantum services? Can authority be revoked in real time when an automated workflow crosses an approved boundary violating trust and authorization?
More critically: Are we protecting today’s information against tomorrow’s cryptographic capabilities?
That question is already implementable today. NIST recently finalized its first three post-quantum cryptography standards (FIPS 203, 204 and 205 in 2024) and encourages enterprises to begin transitioning.
What this means now is that enterprises do not need a cryptographically relevant quantum resource in production before quantum becomes a board-level architecture issue.
That is the distinction between prediction, foresight and preparedness.
Go from cloud-native to compute-agnostic.
Over decades of technology transformation, I have witnessed organizations migrate from mainframes to distributed computing, virtualization, cloud, multicloud and AI-accelerated infrastructure. Each transformation assumed a familiar mistake: assuming the next computing model replaces everything before it.
Quantum should not become another version of that false assumption.
The future enterprise may need to be more compute-agnostic, meaning the ability to deploy workloads on resources that enables the best combination of performance, cost, resilience, throughput, security and business outcome.
Future applications will consume compute as fully governed, ecosystem-agnostic services, and not bind to a processor, cloud, AI model or quantum platform. APIs, data, identity, observability, FinOps and security become the connective tissue.
For investors and founders, business value extends beyond quantum hardware to middleware, orchestration, algorithms, security, developer platforms, industry applications and assurance.
Convergence is not just a research agenda; it is an operating-model imperative. Architecture, cybersecurity, risk, finance, procurement and leadership need shared decision rights to evaluate enterprise advantage. A breakthrough creates no enterprise advantage until it is economically viable, securely integrated, observable and governed at runtime.
What should enterprises do now?
I would not recommend an enterprise-wide “quantum transformation.” Rather, I would build optionality:
1. Identify computationally challenging business problems rather than searching for reasons to use quantum.
2. Establish focused hybrid use cases against measurable classical baselines.
3. Build crypto-agility and begin planning post-quantum migration.
4. Strengthen AI/agentic AI governance because greater capability without bounded authority can only amplify its consequence.
5. Design architectures that can introduce emerging compute resources without rebuilding the enterprise around them.
The strategic objective is not quantum readiness in isolation. It is convergence readiness.
Do not migrate your enterprise to quantum yet. Architect your enterprise so that quantum becomes just another governed computational resource when demonstrable advantage justifies it.
The next compute divide may not separate enterprises that possess quantum computers from those that do not. It may separate institutions that still govern each technology as a separate world from those that can safely orchestrate converging forms of intelligence.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
>> Home