Machine Trust in the First Month: What Should a Business Do Now the Rules Are in Force?
The first month of machine trust is about making AI use visible and closing the most urgent gaps. Because the rules are already in force, this month is not a preparation period but a correction period. Instead of a big project, small but correct steps are taken. 🗓️
This page is the first step of the machine trust horizon series. It covers what is realistic in the first month, which rule to address first, a four-week task list, common first-month mistakes and what is rising on this horizon. Forecasts are labelled as forecasts. This article is for information only and is not legal advice; consult a lawyer about your own situation.
What is realistic in the first month?
BU BÖLÜMÜN ÖZETİ
- An inventory
- Quick fixes
- A responsible person
A month is not enough to build a full compliance arrangement, but it is enough to close the most visible risks and plan the months ahead.

An inventory
The first job is a list of where AI is used: content and image production, advertising, chatbots, call handling and agents. Each row records the tool, the person responsible and where the content is published. Labelling done before the inventory forgets some places and marks others needlessly. The list can be started in a single afternoon and refined over the following weeks. 📋
Quick fixes
The most visible gaps from the inventory are closed in the first month: a customer-facing chatbot saying it is AI, an AI character indistinguishable from a human being disclosed clearly in ads, and content imitating a real person being removed. These three fixes are usually decisions rather than technical work. Small as they look, most complaints and misunderstandings come from exactly these three points.
A responsible person
By the end of the first month, someone should be responsible for machine trust. That person updates the inventory, writes down labelling decisions and follows the rules. Compliance without an owner is forgotten in the first busy week. One owner with a written role is the soundest start for a small business, and it costs nothing but a decision. 🔑
Which rule comes first?
BU BÖLÜMÜN ÖZETİ
- AI in advertising
- Content that reaches Europe
- The 2 December date
Three groups of rules are in force at once. The order depends on the business’s risk.
AI in advertising
In Türkiye, the AI provisions of the advertising regulation have applied since 1 August 2026. For a business that uses AI-generated human-like characters, voices or images in its ads, this comes first. Promoting a product with an AI copy of a real person is prohibited outright. The ad archive is reviewed once with this in mind, including ads still live after a campaign has ended. 📢
Content that reaches Europe
If a chatbot, deepfake content or AI-generated public-interest text reaches people in Europe, Article 50 comes into play. The rule is summarised in EU AI Act Article 50 summary. A business selling to Europe should run this check in the first month rather than wait for a question from a partner or customer.
The 2 December date
Under the European rule, the period for the machine-readable marking duty of generative systems already on the market before 2 August ends on 2 December 2026. That duty belongs to the provider offering the tool. Asking the provider of the tools you use about its marking plan is a smart first-month step, and a written answer is worth keeping. ⏰
A four-week task list
BU BÖLÜMÜN ÖZETİ
- Week 1: the inventory
- Week 2: labels and notices
- Weeks 3 and 4: tools and logs
The sequence below is enough for a small business to get through its first month in order. Each week builds on the one before.

Week 1: the inventory
Every place AI is used is listed, with the tool, the person responsible and where the content is published. Everyone on the team is asked “in which tasks do you use AI?”; most use began informally. Hidden use is the most commonly missed part of the inventory, and it is often where the riskiest data ends up. 🔍
Week 2: labels and notices
Based on the inventory, the team decides which content is labelled and how the chatbot gives its notice. The decision is written down and applied. Label wording should be plain: readers should understand at a glance what they are seeing. How a full labelling system works is covered in setting up an AI content labelling system.
Weeks 3 and 4: tools and logs
The team tests whether the tools in use add provenance information and whether the publishing chain keeps it. If agents are used, check that action logging is switched on. At the end of the fourth week, a short status note is written; it feeds the plan for the third month and becomes the baseline for the first internal audit. 🛠️
Which mistakes happen in the first month?
BU BÖLÜMÜN ÖZETİ
- Labelling everything
- Labelling nothing
- Stripping provenance
First-month mistakes usually come from overreacting or from not reacting at all.
Labelling everything
The European rule does not require a label for every AI touch. A label placed for routine help such as spell-checking overshadows the labels that really matter. A label that appears everywhere loses its meaning for readers. Scope is set by a written policy, and the policy is applied consistently. 🏷️
Labelling nothing
The assumption “this doesn’t concern us” is risky, especially for a business that uses AI characters in its ads. Türkiye’s regulation binds every advertiser, not just influencers. Checking instead of assuming is the rule of the first month. Even if you conclude you are out of scope, writing down why makes life easier when someone asks later.
Stripping provenance
When an image is edited, compressed or re-saved, provenance information in the file can be removed without anyone noticing. Testing the publishing chain once prevents that loss. A stripped marker cannot be found later when proof is needed, and rebuilding it after the fact is rarely possible. 🧾
What is rising on this horizon?
BU BÖLÜMÜN ÖZETİ
- Today: 2 December is approaching
- Forecast: verification becomes visible
- Forecast and log
Most of what will happen on a one-month horizon has already begun. The first item is today’s status; the next two are forecasts.

Today: 2 December is approaching
The marking transition under the European rule ends on 2 December 2026. After that date, generative AI tools serving Europe are expected to mark their outputs. Asking about marking support when choosing tools should become a habit from today. A tool that cannot answer the question clearly is a risk worth noting in the inventory. 📅
Forecast: verification becomes visible
Forecast: in the coming months, search engines and browsers will show content provenance to users more visibly; Google announced in May 2026 that it would bring C2PA verification to Search and Chrome. Content that carries provenance information will gain a trust advantage during this period, while stripped content will stand out.
Forecast and log
Forecast: businesses that complete their inventory in the first month will adapt at lower cost as the details of the rules become clearer. Update log. 12 November 2026: first publication. The forecasts on this page are re-read when they fall due; a forecast that turns out wrong is not deleted but corrected here. The next step in this series is machine trust: three months.
How does the first month work with Adapte Digital?
BU BÖLÜMÜN ÖZETİ
- A single delivery package
- What is delivered
- Where to start
We do the first month’s work as a delivered job: the inventory, the quick-fix list and the labelling decisions are prepared and handed over to you.
A single delivery package
Adapte Digital does not charge a monthly management fee. Work starts with a single delivery package whose scope is written down in advance: 36,000 TRY + VAT. If the work grows, the new need is defined as a separate add-on package, priced by what will be delivered rather than by the hours spent. 📦
What is delivered
The output is an AI usage inventory, a prioritised fix list and written labelling decisions. A deeper look at internal AI use is described in business AI usage audit.
Where to start
Which packages are open, and at what price, is shown on our campaign pages, item by item. 🧭
Quick Summary
- The first month is for an inventory and quick fixes.
- AI use in ads has been regulated in Türkiye since 1 August 2026.
- The EU marking transition ends on 2 December 2026.
- Labelling everything and labelling nothing are the two extreme mistakes.
- Forecasts are revisited in the update log.
Short Glossary
- AI inventory
- A list of where a business uses AI, with the tool, the person responsible and where content is published.
- Transition period
- Extra time granted in a law for complying with an obligation.
- Hidden use
- AI tools employees use in work processes without formal approval.
Frequently Asked Questions
Next Step
This week, ask your team in which tasks they use AI and gather the answers in one list. Our campaign pages list the delivery packages that cover this work.
Sources: EUR-Lex — Regulation (EU) 2024/1689 (Artificial Intelligence Act) · Alomaliye — Reklam Yönetmeliğinde Kapsamlı Değişiklik
Updated: November 2026
Sık Sorulan Sorular
The inventory of where AI is used. Without it, you cannot know which rules bind you.
Add a clear notice so customers understand they are talking to AI; if you serve users in Europe, this may fall under Article 50.
The date concerns providers’ marking duty. Asking the provider of your tools about its plan is a sufficient first step.
