The price sheet for a frontier language model lists a flat rate per million tokens. It does not list the compounding cost effect that emerges when those tokens are assembled into long-context requests. This distinction — between the unit price of a token and the system cost of a context window — is where the most dangerous budget surprises of 2026 are occurring. A healthcare organisation recently discovered this gap at the scale of a $6 million overrun in a production retrieval-augmented generation (RAG) pipeline. Understanding what happened requires reasoning about the physics of context before you can reason about the economics.

Frontier coding assistants are capable, but they arrive with constraint layers baked in. For security researchers, low-level systems work, or any task touching edge-case logic, those constraints become a productivity tax rather than a safety net. The model declines; the sprint stalls.

The EU Data Act introduces rules on access to and use of data, contractual fairness, switching between data processing services, and interoperability in several contexts (including Internet of Things and B2B data sharing scenarios per its scope). It is explicitly aimed at reducing lock‑in and unfair contractual imbalance—not identical to GDPR, but complementary when you negotiate cloud and platform contracts.

In early 2026, Uber’s CTO Praveen Neppalli Naga disclosed something that should have landed harder than it did: his engineering organisation had burned through its entire annual AI budget before the calendar reached May. The culprit was not a rogue data science project or an experimental LLM fine-tune. It was a coding assistant — specifically, Anthropic’s Claude Code — running at roughly $2,000 per engineer per month across a team large enough to make that number catastrophic at scale. This is the first instalment of an eight-part series on the 2026 AI Token Economy: what it costs, why costs escape, and how to govern them before your own budget evaporates.

The Cyber Resilience Act (CRA) targets products with digital elements placed on the EU market: secure by design, vulnerability handling, and transparency obligations that flow through manufacturers, importers, and distributors. If you ship hardware or software as a product—or embed connectivity in what you sell—CRA logic eventually touches how you build, update, and disclose flaws.

The demo was never lying: a few dollars of API spend can draft what might have taken a human afternoon. In production you run two ledgers. One is the vendor bill—frontier models are still on the order of dollars per million input tokens and more for output (OpenAI API pricing). The other is retries, review, security fixes, and escalations when “almost right” ships.

The EU AI Act regulates how AI systems are placed on the market and used, with escalating obligations for high‑risk categories, transparency for certain general‑purpose and consumer‑facing cases, and governance expectations that land on deployers as well as providers. Open‑source weights or code do not automatically exempt a real‑world deployment from duties once the system is part of a product or business process.

NIS2 raises the floor for cyber risk management, supply‑chain security, incident handling, and accountability across essential and important entities in Member States’ transpositions. Open source can strengthen security when you control patching and architecture; it weakens it when “community maintenance” becomes an excuse for drift.

The Digital Operational Resilience Act (DORA) is not a licensing discussion; it is an ICT risk and third‑party regime for the financial sector. Whether your database is proprietary or open source, you must prove you can manage and recover critical functions when suppliers, software, or infrastructure fail.

The GDPR does not mandate open source. It mandates accountability: lawful basis, purpose limitation, data minimisation, security of processing, subprocessors, transfers, breach notification, and the rights of data subjects. When you migrate workloads to open‑source platforms or self‑hosted stacks, supervisors still ask one question: can you show how controls are implemented and who is responsible?

Proprietary stacks are not “bad,” but they concentrate risk. When critical workloads, encryption keys, audit logs, and AI inference paths all depend on opaque roadmaps and a single commercial trajectory, you inherit someone else’s priorities: price changes, feature deprecation, regional product splits, and incident timelines you do not control.

Over this series we moved from one core idea to its full implication: once AI is connected to business systems, it becomes an attack surface with real operational consequences.