Know what company data is going into artificial intelligence.
Our analysts install an AI gateway in monitor mode. It records usage and flags sensitive data such as Brazilian taxpayer IDs (CPF), card numbers and passwords. At the end, you receive the report and the use policy.
What secure AI use is
When an employee pastes a customer spreadsheet into ChatGPT, the data leaves the company with no record. Secure AI use means knowing which artificial intelligence tools your team uses and what is sent to them. CyberArmor monitors that use for 30 days, without interrupting work, and delivers a report and a usage policy.
An AI gateway sits between the employee and the tool. In monitor mode, it records usage and flags personal data, credentials and sensitive attachments.
The services in this practice
The assessment is the starting point. It ends with a proposal for ongoing governance.
- 30 daysGood place to start
AI use assessment
Our analysts turn on the AI gateway in monitor mode for 30 days. It observes usage and blocks nothing. At the end, you receive the report of what was seen and the company’s AI use policy.
Fixed price quoted within 1 business day
- Ongoing
Ongoing AI governance
An assigned analyst follows AI use every month, adjusts the policy rules and sends a monthly report. Blocking directly in the browser is not available yet and joins the service when it launches.
Want a first picture before we talk?
The free self-assessment has 12 questions and shows your company’s AI use maturity level and its 3 priorities. It runs in your browser and requires no sign-up.
Take the AI use self-assessmentWhat you receive
Every capture is recorded on the platform, with a date and with the sensitive data masked.
- An AI use report: which tools, in which departments and how often.
- Sensitive data found in submitted text and attachments, such as CPF, card numbers, passwords and access keys.
- An AI use policy written for your company, ready for board approval.
- Recommendations in order of priority to reduce risk.
How the work happens
Assessment
The analyst learns how the company uses AI today and which data matters most to the business.
You receive: a map of the departments and the data that matter.
Installation in monitor mode
The analyst installs the AI gateway and turns on monitor mode. People keep working as usual.
You receive: confirmation that captures are reaching the platform.
30 days of observation
The gateway records usage and inspects submitted text and attachments. The analyst reviews captures every week.
You receive: the first findings, with sensitive data masked.
Use policy
The analyst writes the AI use policy based on what was observed.
You receive: the policy, ready for approval.
Executive report
The assessment closes with the board report and the proposal for ongoing governance.
You receive: the report and dated evidence.
Who does the work
A team of experienced, certified security and data protection analysts. Each project has an assigned analyst who talks directly to your team.
- The assigned analyst follows the project from installation to report.
- Every capture and finding is recorded on the platform, with a date.
- Text sent to the AI is not stored: the platform shows masked evidence.
The résumés and certifications of the assigned team are presented in the proposal.
Why CyberArmor
What you get in every contract, from day one.
- An assigned analyst. Each project has an analyst who answers for the work from start to finish.
- Dated evidence. Every finding, fix and piece of evidence is recorded on our platform, with the date it happened.
- A deliverable at every step. You know what you receive in each of the five steps of the method.
- A report for the board. The executive report explains risk in business terms.
On the platform
The AI content capture screen shows whether the sensor is reporting, how many captures arrived and when the first and last happened.
- The AI content capture screen, in the governance area.
- Sensor status: whether captures are actually arriving.
- How many captures there were, and when the first and last happened.
How to hire us
Every contract starts with a fixed project. It ends with a proposal for an ongoing service, if that makes sense for you.
- Fixed project
- Scope, timeline and price agreed before we start. It takes 2 to 4 weeks and ends with the report and the plan.
- Ongoing service
- An assigned analyst looks after your environment every month, with a monthly report and evidence on the platform.
Frequently asked questions
What is an AI gateway?
It is a checkpoint between your people and AI tools. It records usage and inspects what is sent before the data reaches the tool.
Does monitor mode block anything?
No. It only observes and records. That way the company learns its real usage before deciding on any rule.
What data does the inspection find?
Brazilian data such as CPF (taxpayer ID) and card numbers, plus passwords and system access keys. Inspection covers submitted text and attachments.
Which AI tools are covered in the assessment?
The specialist confirms the list of tools covered for your case in the first conversation.
What about the LGPD?
The LGPD is Brazil’s General Data Protection Law. It requires companies to know where the personal data they handle goes. The report shows which personal data was sent to AI tools, and the policy defines what is allowed.
How can a company control the use of ChatGPT?
Start by measuring. Before banning or allowing it, the company needs to know who uses ChatGPT and other AI tools and what data they send. With that picture, it can write a usage policy the team is able to follow. CyberArmor's AI use assessment takes that measurement in 30 days, without blocking work.
What does the AI assessment report show?
It shows which AI tools the team used during the 30 days and what kinds of sensitive data were sent to them: personal data such as CPF numbers, card numbers, passwords and access keys, including inside attachments. The usage policy is written from that picture, stating what is allowed and what is not.
Talk to a specialist for 30 minutes.
Tell us what your company needs. You leave the call with three priorities for your environment.
Prefer to write? Use the email contato@cyberarmor.com.br.