IN DEVELOPMENT · AI-IDEENSCHMIEDE

AI acts.
You decide.

AI protection against AI attacks

We are developing AI-Certain as an independent control layer between AI assistants and the data, applications and APIs they want to access.

How AI-Certain is designedExplore the demo

This interface is a demo. Technical protection will depend on the security core and verified integrations.

ACTION CHECKConcept example
An assistant requests an actionReview tomorrow’s reservations
AI-CERTAIN

Check first.
Then execute.

TaskPermissionsData & destination
AllowWithin scope
HoldApproval needed
BlockBoundary breached
Enforced boundaries outside the AI model.
01 / OUR APPROACH

Responsibility
stays with people.

An AI assistant can analyse information and propose actions. AI-Certain is designed to check whether a requested action fits the task, which permissions apply and what data would leave the system.

Our security approach combines technical boundaries with decisions you can understand. Critical actions are held or blocked. Approval stays with the user wherever their rules require it.

Our principleAn AI must not expand its own permissions or disable its own safeguards.
02 / PLANNED FEATURES

A framework for
your AI assistants.

These building blocks guide our development.

01

Your rules

Define which data may be used, which destinations are permitted and which actions must remain blocked.

02

Tasks & permissions

Give each assistant a specific task with appropriate permissions, a time limit and a cost limit.

03

Multiple AI models

Choose models through supported integrations and divide tasks between assistants with separate roles.

04

Specific approvals

Review the exact action and its effects. Each approval is intended to apply once to that specific action.

05

Emergency stop

Stop one or all assistants using a visible button or your personal stop phrase through your authenticated control channel.

06

A clear activity log

Track what was requested, allowed, held or blocked, with clear reasons.

03 / HOW IT IS DESIGNED TO WORK

One task.
Three clear boundaries.

Example task: “Review tomorrow’s reservations.”

Choose an action to see the intended decision.

ALLOWED

Reading is part of the task.

The assistant may read the required reservations. Access remains limited to the defined scope.

Rule: reading the required data is allowed
CONCEPT EXAMPLE · NO REAL SYSTEM ACTION
04 / OUR WORK

Building control
step by step.

AI-Certain is being developed at AI-Ideenschmiede. We combine human experience with AI as a tool and review the results of our work.

Our goal is an understandable security solution with technically enforced rules. A second AI can provide additional insights. Binding access limits are intended to be enforced by the independent security core.

More about AI-Ideenschmiede
  1. DEVELOPED

    Concept & interface

    Rules, permissions, models, approvals and emergency stop controls have been developed as a concept and interactive demo.

  2. NEXT STEP

    Security core & first integration

    Check actions on the server and execute them under control. Permissions and stop controls must be enforced independently of the assistant.

  3. PLANNED

    Test, measure, improve

    Test permitted and prohibited actions, manipulation attempts and failures. Use those results to build further integrations.

05 / ONE CORE, SUITABLE INTEGRATIONS

AI-Certain and AILENDO Guard.

STANDALONE SOLUTION · PLANNED

AI-Certain

The shared security core and management interface for connected AI assistants and systems.

INTEGRATION · PLANNED

AILENDO Guard

The same approach, with rules and interfaces for AI features within AILENDO. The user stays in control.

06 / FREQUENTLY ASKED QUESTIONS

What you should know today.

Is AI-Certain already a finished security product?

We are still in development. The integrated interface demonstrates the controls using sample data. It is not connected to an AI model or production system and does not block real attacks.

What does “AI protection against AI attacks” mean?

We focus on controlled AI actions: manipulated content must not give an assistant extra permissions. AI-Certain is designed to enforce the user’s boundaries independently of the model. This does not require identifying an attack as “AI-generated”.

Can I use several AI models at the same time?

That is planned. Assistants can be assigned different models and roles. Communication between them is intended to follow the same data and permission rules. Supported models will be determined through verified integrations.

Can the emergency stop undo everything?

The planned stop blocks new actions. Cancelling an action already in progress depends on the relevant interface. Data already sent and actions already completed cannot be reversed by the stop.

Which systems will AI-Certain protect?

Protection is planned for actions routed through controlled integrations. Each integration requires its own technical verification. Websites, servers and user accounts also require standard security measures against general attacks.

When can I buy AI-Certain?

A release date and price have not yet been set. Distribution is planned through Ailorium. The demo is free to explore.

07 / INTERACTIVE DEMO

Try the controls.

Choose models, draft rules and tasks, review an approval and test the emergency stop. Everything runs locally with sample data. Settings are temporary and do not enforce real security rules.

Open demo in its own view
Control for your AI assistants
Demo: monitoring active
Workspace · AI-Ideenschmiede

Security overview

Interactive concept · Sample data · No connected systems
CONTROLLED ACCESS · DEMO

One action needs your decision.

Your boundaries are defined. One reservation change is on hold; an external data export is blocked. All values below are simulated sample states.

ACTIVE ASSISTANTS
2
Separate roles and permissions
PENDING APPROVALS
1
Waiting for your decision
BLOCKED ACTIONS
1
No data transferred
ACTIVE RULES
3
No expansion by assistants

Assistants & tasks

AAILENDO assistantActive
Model A · Cloud · Sample assignment
Task: review reservations
READ + REQUEST CHANGES
RReview assistantActive
Model B · Cloud · Sample assignment
Task: review change drafts
READ ONLY

Recent security events

Blocked
Customer data export blocked

Destination not permitted · Outside task scope

On hold
Pickup time change on hold

Waiting for one-time user approval

Allowed
Reservations read

Within task scope · Read permissions checked

DECISION REQUIRED
Approval required

Change pickup time

Reservation R-1042 · 5 October 2026

10:00 → 11:00
Data
Pickup time only
Recipient
No external recipient
Approval
Once, for this change

Enforced boundaries

01
Customer data stays in the system

External recipients are not permitted.

02
Permission changes are blocked

Assistants cannot expand their own access.

03
Changes require user approval

A specific action, checked before execution.

Sample state · No AI models or systems connected
Developer: AI-Ideenschmiede · Distribution: AiloriumAI-Certain · Demo 0.7.0
PEOPLE + AI + CLEAR BOUNDARIES

Take a closer look
at our work.

Explore the interface and see how rules, approvals and stop controls are designed to work together.

Try the demoVisit Ailorium
AI-Certain · Legal information

Legal notice

Provider

AI-Ideenschmiede
Owner: Jörg Schwarz
Am Pulverschoppen 11
59071 Hamm
Germany

Contact

Email: shop@ailorium.de

About this website

This website introduces the AI-Certain development project. The demo presents an interface with sample data. It does not implement a live AI integration or actual security protection.

Copyright

© 2026 AI-Ideenschmiede · Jörg Schwarz. Rights to our own copyright-protected content are reserved. Uses permitted by law remain unaffected.

AI-Certain · Legal information

Privacy policy

This policy covers the information website ai-certain.com and its integrated interface demo, including demo.html and the Guard Lab at risk.php. Separate information will be required for any future production AI integration.

Data controller

Jörg Schwarz · AI-Ideenschmiede
Am Pulverschoppen 11, 59071 Hamm, Germany
Email: shop@ailorium.de

Website delivery and hosting

The website is intended to be hosted by STRATO. When you visit, technically necessary connection data is transmitted to the hosting provider, such as your IP address, the requested page, the time and browser information. This data may be processed in server logs to deliver the website and secure its operation. The legal basis is Article 6(1)(f) GDPR; our legitimate interest is reliable and secure operation. The scope and retention of server logs depend on the actual hosting configuration and agreements with STRATO.

Cookies, analytics and external resources

The informational pages and browser-only interface demo do not set cookies or use persistent browser storage. Guard Lab (risk.php) uses a necessary session cookie to associate test requests, rules and decisions with your browser. It stores this test state in a PHP session on the hosting server, with an application inactivity limit of 30 minutes. Physical cleanup of expired session files depends on the hosting configuration. Up to 50 requests and 100 log entries are held per test session. No customer data or payments are sent to external systems. They do not include analytics or advertising services, external fonts or externally loaded libraries.

Demo and your inputs

The interface demo in index.html and demo.html runs in your browser. The separate Guard Lab sends test selections and decisions to the hosting server; it has no AI-provider connection. The browser-only interface demo does not send your settings or inputs to a server or AI provider, and they are lost when you reload the page. Please use sample data. Requests for the demo files are processed by the hosting provider like other page requests.

Agent access test

The optional agent-test.php test uses a necessary operator session cookie and a short-lived separate agent credential. Synthetic test actions are sent to this hosting server over HTTP secured by HTTPS. The server stores run identifiers, agent and task identifiers, a credential hash, permission scopes, timestamps, decisions and adapter invocation counters in private files outside the public web directory. Credentials expire after five minutes. Old run files are removed on a subsequent test start after they are more than 24 hours old. Removal is not guaranteed at exactly 24 hours if no new test starts. The operator session has a 30-minute inactivity limit. These test records are not a permanent audit archive. There is no connection to an AI provider or real customer system.

Contact by email

If you contact us by email, we process your message and contact details to handle your enquiry. The legal basis is Article 6(1)(b) GDPR for contract-related enquiries, or Article 6(1)(f) GDPR based on our legitimate interest in handling other incoming enquiries. The data is deleted when the enquiry has been handled, unless statutory retention obligations apply.

External links

Links include AI-Ideenschmiede and Ailorium. If you follow a link, the destination website’s data processing practices and privacy policy apply.

Your rights

Subject to the conditions provided by law, you have rights of access, rectification, erasure, restriction of processing and data portability. You may object to processing based on legitimate interests on grounds relating to your particular situation. You may lodge a complaint with a data protection supervisory authority, particularly where you live, work or believe an infringement has occurred.

Last updated: 17 October 2026.

AI-Certain · Legal information

Copyright & use

AI-Certain is a development project by AI-Ideenschmiede, owned by Jörg Schwarz.

Protected content

© 2026 AI-Ideenschmiede · Jörg Schwarz. To the extent that our own text, design, documentation and software are copyright-protected and we hold the rights, those rights are reserved. Uses beyond those permitted by law require the consent of the relevant rights holder. Third-party rights remain unaffected.

Demo and sharing

You may explore the public demo through its intended interface. This does not grant a general licence to incorporate protected elements into your own products. You may share links to our website. Uses permitted by law remain unaffected.

Idea, name and technical development

This notice does not grant exclusive protection for the general idea of a security control layer for AI agents. Trademark and patent protection have separate requirements; this website does not claim a registered trademark or a granted patent for AI-Certain.

Collaboration

Publishing this concept does not establish a confidentiality agreement. Non-public technical details will only be disclosed during a collaboration after a separate agreement has been made. For enquiries about cooperation and usage rights: shop@ailorium.de.