IBM + OpenAI: How Consulting Giants Are Cornering Enterprise AI

What Actually Happened on August 13, 2026

On August 13, 2026, IBM and OpenAI announced a partnership that looks like just another technology deal on the surface — and marks a strategic shift on closer inspection. IBM Consulting is building a dedicated OpenAI practice, training tens of thousands of consultants on GPT-5.6, Codex, and ChatGPT Work, and embedding these models into industry-specific solutions for financial services, government, telecom, and retail — integrated into IBM's own Consulting Advantage platform. No financial terms were disclosed; as an operational resource, IBM is deploying so-called Forward Deployed Engineers who work directly alongside clients on implementation.

Why This Isn't an Ordinary Software Partnership

The real story isn't the scale of the training — it's the business model shift behind it. With this move, OpenAI is stepping away from the classic SaaS approach — API access, self-service, software-driven scale — and instead scaling through consulting integrators. The model is no longer 'buy access,' but 'buy implementation through a partner who controls the implementation.' Whoever controls implementation controls the client relationship — and, over time, pricing power.

This isn't an isolated move. It's a pattern showing up across the industry simultaneously:

• Anthropic has launched its own AI-native enterprise services, backed in part by Blackstone, Hellman & Friedman, and Goldman Sachs.
• OpenAI itself is building a 'Deployment Company,' backed by more than $4 billion from TPG, selling implementation rather than pure software.
• Coforge launched Momentuum AI, a Forward Deployed Engineer model with outcome-based pricing.
• Kyndryl is positioning 'Agentic Modernization' explicitly as services-as-software.

The large consultancies are restructuring accordingly. Capgemini, for instance, is organizing its offering around five AI value pools with an 'agentic control plane' as the central orchestration layer — further evidence that control over AI orchestration, not any single model, has become the real competitive battlefield.

What This Means for Swiss SMEs

For Swiss decision-makers, this creates an uncomfortable but clear starting position. As the major consulting houses — IBM, Accenture, Deloitte, Infosys, TCS — systematically lock in enterprise AI deployment expertise before companies can build it themselves, negotiating power shifts. External help either becomes more expensive and concentrated among fewer large providers, or companies need to start building internal capability now, while the market hasn't fully consolidated yet.

The Capability Clock Is Ticking Faster Than Many Assume

The data on where things stand today is clear: according to the OECD D4SME programme 2026, 34 percent of Swiss SMEs currently use any form of AI, 76 percent of those remain at a novice level, and only 3.6 percent qualify as 'champions' with mature processes. Companies that deliberately choose external expertise over building it themselves achieve up to three times higher success rates — provided the choice is made deliberately, not under time pressure.

Three Courses of Action for Swiss Decision-Makers

Three realistic paths emerge from the current market shift, and they aren't mutually exclusive:

1. Decide now, don't wait: the build-versus-buy question can't be deferred indefinitely without weakening your negotiating position. A structured evaluation of in-house development versus an external partner belongs on the leadership agenda.
2. Use compliance as a selection filter: the Digital Omnibus of June 29, 2026 pushed the EU AI Act's high-risk obligations to December 2, 2027 — buying more time, but not reducing the need to select vendors based on compliance maturity rather than price alone.
3. Actively evaluate Swiss alternatives: alongside the global consulting giants, an independent Swiss ecosystem is emerging — from sovereign models like Apertus to specialized providers for SME use cases such as phone agents or customer service. These options reduce dependency on a single global provider and deserve a place in every evaluation.

What Matters Now

The IBM-OpenAI partnership isn't an isolated event — it's a signal. The windows in which negotiating power and market access can still be redistributed are closing. For Swiss SMEs, this doesn't mean acting rashly — it means deciding deliberately, with whom and how to build AI capability, before that decision is made for them by a narrow oligopoly of global consultancies.