Adopting AI is a project. Running it is a full-time job.
The expensive part of AI automation is not the pilot. It is the 24 months that follow — and it appears in no quote.
The prototype works. The assistant answers invoicing questions, the agent classifies documents, the model pulls deadlines out of contracts. Management is impressed, IT is proud, the business case adds up on paper.
Then six months pass.
The model provider announces a retirement date for the version you use. A department introduces a new document format the extractor has never seen. Someone notices accuracy has been slipping for weeks — but nobody can say since when, because there is no baseline. Legal asks for the AI Act documentation. And the colleague who built the whole thing is now a team lead with no time left.
This is not an edge case. This is the normal case.
What the pilot costs — and what comes back every year after it
In person-days, mid-sized company running two to three production use cases
AI is not a software rollout
With conventional software the arithmetic is simple: install once, then it runs. Updates come from the vendor, maintenance is a predictable percentage of licence cost, and if nobody touches it, the system behaves the same in two years as it does today.
AI systems behave differently. They decay even when nobody touches them — because everything around them moves.
- The models moveProviders retire versions, change response behaviour, shift safety boundaries. A prompt that reliably returned structured JSON in January produces prose with a preamble in June. Nobody notices automatically — unless someone built regression tests and keeps them current.
- The data movesNew contract templates, renamed ERP fields, a different scanning process in the mailroom. Any of these can quietly degrade a retrieval pipeline. Not break it — degrade it. That is the more dangerous version, because the system keeps answering, just worse.
- The regulation movesThe EU AI Act transparency obligations have applied since 2 August 2026. Germany's NIS2 implementation act has applied since December 2025 with no transition period, including personal liability for management. Neither asks for a one-off action; both ask for ongoing evidence.
- The attackers moveInjected instructions inside uploaded documents, data leaking through over-broad tool permissions, model endpoints without limits. An AI system with access to internal systems is a new attack surface — and one most companies have no hardening playbook for.
An unmaintained AI system does not break. It quietly gets worse.
Schematic view of the typical quality curve after go-live
The regulatory clock runs alongside
Anyone running AI runs it inside a legal framework that changes faster than their own roadmap. The Digital Omnibus pushed the high-risk obligations back in July 2026 — that buys preparation time, but removes no obligation. And the obligations already in force are the ones mid-sized companies most often miss.
What applies now and what is still coming
EU AI Act and NIS2 in order of deadline, not to scale
What this actually costs
The breakdown below assumes a realistic mid-market scenario: 150 to 500 employees, two to three production use cases — document extraction, an internal knowledge assistant, a compliance check — operated in-house.
The basis is €800 per person-day for internal work, meaning fully loaded cost rather than gross salary, and €1,400 for external contracting. The range in the cost column reflects exactly those two ends.
Twelve cost blocks that recur every year after go-live
Annual range in thousands of euros
| Cost block | What sits behind it | Effort p. a. | Cost p. a. |
|---|---|---|---|
| Model and provider changes | Tracking retirement dates, testing version changes, fallback routing, comparison runs | 15–30 PD | €12,000–42,000 |
| Prompt and retrieval upkeep | Updating prompts, adjusting chunking, re-indexing, onboarding new document types | 25–45 PD | €20,000–63,000 |
| Quality measurement | Building and maintaining a reference set, regression tests, error rates, drift monitoring | 20–35 PD | €16,000–49,000 |
| Regulatory work and evidence | AI Act Art. 4 and Art. 50, Annex III preparation, ISO/IEC 42001, audit documentation | 15–30 PD | €12,000–42,000 |
| Data protection | Impact assessments, processing agreements, technical measures, deletion policies, transfers | 8–15 PD | €6,400–21,000 |
| IT security | Hardening against injected input, secrets management, permissions, penetration testing, patching | 15–25 PD | €12,000–35,000 |
| Operations and incidents | Logging, alerting, on-call, incident handling, restore testing | 20–40 PD | €16,000–56,000 |
| Infrastructure | Compute and hosting, vector database, observability, staging environment | Direct cost | €9,000–60,000 |
| Model usage under load | Usage-based billing or running your own models under sustained load | Direct cost | €6,000–48,000 |
| Training and AI literacy | Obligation under AI Act Art. 4, role-specific training, onboarding new staff | 8–15 PD | €6,400–21,000 |
| Rework and technical debt | Framework migrations, architecture changes, paying down prototype shortcuts | 20–40 PD | €16,000–56,000 |
| Knowledge continuity | Hiring and onboarding, documentation, cover arrangements | 5–15 PD | €4,000–21,000 |
| Total | 151–290 PD | €135,000–515,000 |
Basis: €800 per internal person-day (fully loaded), €1,400 external. Direct costs estimated separately.
Running three to five use cases will not triple the person-days, but it will not keep them flat either. Realistically the factor is 1.8 to 2.2.
Where those days go is just as telling. Half of them go not into compliance or security, but into the plain work of catching the technology up with changes that happened somewhere else.
Where the person-days go
Midpoint of the range, roughly 220 person-days per year
The line items that appear in no table
- The bus factorIn most mid-sized companies building AI in-house there is exactly one person who understands why the system is built the way it is. When that person resigns, goes on leave or gets promoted, what begins is not maintenance but archaeology. Rebuilding typically costs 40 to 60 percent of the original development effort.
- The opportunity costThe person spending 200 person-days a year maintaining AI systems is usually the best engineer or architect in the building. What they would have produced in product or process work over the same period appears on no cost centre — and is often the most expensive item in the whole calculation.
- The evidence gapReconstructing model versions, prompt states, test results and decision records once the audit has started costs a multiple of what continuous logging would have cost. Under section 65 of the German BSIG, late NIS2 registration alone can carry a fine of up to €500,000; the AI Act penalty range reaches €15 million or 3 percent of global annual turnover in the standard case.
- Write-offs on technology shiftsWhat was built in 2024 on the framework of the day is often not the obvious architecture in 2026. Building in-house means paying for those shifts yourself — and deciding for yourself when to follow them.
In-house or specialist provider
The honest comparison, including the rows where in-house wins.
| Criterion | Built in-house | Specialist provider |
|---|---|---|
| Time to production | 4 to 9 months | 4 to 10 weeks |
| Ongoing operating cost | €135,000–515,000 per year | predictable service fee |
| Model updates | your risk, your effort | included in the service |
| Regulatory monitoring | has to be built | part of the product |
| Audit evidence | produced by you | generated continuously |
| Knowledge risk | high, often one person | distributed at the provider |
| Data sovereignty | maximum | depends on hosting model, the key selection criterion |
| Domain control | maximum | high where processes are configurable |
| Handling edge cases | unlimited | bounded by product scope |
| Scaling to new use cases | each one a new project | incremental |
The two highlighted rows are why the answer is not always to outsource.
When building in-house is the right call
This calculation is not an argument against internal development. It is an argument against unexamined internal development. Building it yourself is right when at least two of these four apply.
- The use case is core to the business model and therefore a competitive advantage in itself.
- At least three people in-house have real operating experience with AI systems — not one person with an interest in the topic.
- No market product covers the domain.
- The data situation is specialised enough that a standard product would need more adaptation than a build.
If none or only one applies, in-house is usually the most expensive option — and it only reveals itself as such after 18 months, when unwinding it already hurts.
Three questions before any AI rollout
- Who is named on the org chart as responsible for model changes, quality measurement and audit evidence?And how many days a year are they allocated?
- How would we notice quality dropping?If the answer is “when somebody complains”, there is no measurement.
- What happens if that person leaves?If the answer is uncomfortable, the risk has already been taken.
Our approach
This is exactly where Agentic360 comes in. We run a compliance automation platform for German mid-sized companies in which the operating tasks listed above do not sit with the customer: model changes, quality assurance, regulatory monitoring and evidence generation run on our side.
Two things matter to us. Data sovereignty stays with the customer — the platform runs locally hosted models on German infrastructure, and on-premise and air-gapped installations are part of the product rather than an exception. And decisions stay traceable: rules are evaluated in a deterministic policy engine, not hidden inside a model prompt. Every assessment traces back to the underlying regulation and the processing step that triggered it. That is the difference between “the system says X” and “the system says X because”.
If you are currently working out what your own AI operation costs, or realising you have never done that calculation: talk to us. Even if the answer turns out to be building it yourself — at least let it be a deliberate decision with a budget behind it.
This calculation is not an argument against internal development. It is an argument against unexamined internal development. Even if the answer turns out to be building it yourself — at least let it be a deliberate decision with a budget behind it.
As of August 2026. Regulatory deadlines change; the details reflect the position at publication. This article is not legal advice. The cost figures are estimates based on the stated day rates, not measurements.
- Regulation (EU) 2026/1744 (Digital Omnibus on AI), Official Journal of 24 July 2026, in force since 27 July 2026 — EUR-Lex
- Regulation (EU) 2024/1689 (EU AI Act), Art. 4, Art. 50, Art. 99 — EUR-Lex
- German NIS2 Implementation and Cybersecurity Strengthening Act, in force since 6 December 2025; sections 30, 33, 65 BSIG
- German AI Implementation Act, Bundesrat 10 July 2026 — Bundesnetzagentur as central supervisory authority