Carelog is GDPR-compliant and D-Mark certified. Learn about our security measures for protecting patient data in Denmark.
D-mærket
The Danish label for digital responsibility
D-mærket (the D-seal) is Denmark's labelling scheme for IT security and responsible data use. It is awarded to companies that, based on an annual audit, meet D-mærket's requirements for IT security and responsible data use.
Our data protection controls and GDPR compliance are verified through an independent ISAE 3000 assurance report - third-party documentation that our processes deliver what we promise.
How we protect our users
Our system integrates a range of security measures and protocols to ensure that personal information and health data are handled with the highest level of security.
01
Local storage on your device
All your notes and client information are stored exclusively locally on your own device and encrypted with strong, industry-standard encryption (OS-native encryption on desktop and AES-256-GCM in the browser). Carelog has no access to your data, and no content is stored permanently on our servers. You retain full control of your data at all times.
02
Real-time transcription
Audio from the consultation is transcribed in real-time using a GDPR-compliant speech-to-text API. For real-time speech-to-text processing, chunks of audio are continuously processed in short-term memory, and no audio data is stored. Therefore, there is never, at any point, a complete audio recording of the consultation. Only individual chunks of audio are processed and then immediately discarded. All data processed by Carelog is encrypted for protection.
Based on guidelines from the Danish Council on Ethics
9 Key Findings from Project
1
Finding 1: Data ethics is already part of practice
Ethical considerations are a cornerstone of healthcare professionals' work and influence many decisions - including when it comes to the use of AI.
2
Finding 2: AI is implemented based on many different motives
Efficiency is far from the only driving force. Motivations range from professional curiosity and innovation through organizational modernization to the desire to improve health for patients and citizens.
3
Finding 3: AI provides the most value when linked to a clear need
Complaints procedure
Do you want to complain or object to our processing of personal data or our IT security? See our complaints procedure
Get started with Carelog today
Join thousands of clinicians enjoying stress-free documentation.
Only you have access to your transcriptions and anonymized notes through your two-factor protected user interface. This allows you to review and finalize the notes before transferring them to your medical records system. All transcriptions and notes are automatically and permanently deleted after 5 days. Before this, you also have the option to delete them manually.
04
Your data
Neither Carelog nor third parties use your data to train models or for any other purposes. All information processed is only accessible to you. In other words, no employees at Carelog or other involved data processors have access. To maintain the highest level of security, all data processing takes place within the EU in accordance with the European Union's General Data Protection Regulation (GDPR). Temporary storage of data occurs on dedicated and private Microsoft Azure servers in Sweden.
05
Transparency about security
We prioritize transparency in our security practices. Should a security incident occur that affects your data, we will inform you promptly in accordance with GDPR guidelines. Additionally, we regularly share updates here about new security measures we implement to ensure that our services meet the highest standards.
AI has great potential when used to solve a specific problem. However, in some cases, implementation occurs before the need is clearly defined.
4
Finding 4: Data and model selection is crucial
The development of meaningful AI solutions heavily depends on the choice of data and models.
5
Finding 5: It is complex and resource-intensive to create responsible AI
It is often a major task to navigate the requirements for development, approval, labeling, and documentation.
6
Finding 6: Consent can be difficult in practice
When AI is integrated into healthcare workflows, it can be difficult to obtain individually informed consent for the use of the technology.
7
Finding 7: Uncertainty and errors are central themes in working with AI
There is significant attention to how errors and uncertainty are handled in AI systems.
8
Finding 8: The desire for scaling
Several actors want their AI solutions to be scalable and create value broadly in the health and welfare sector.
9
Finding 9: There is a need for interdisciplinary collaboration
Interdisciplinary collaboration is necessary to succeed in developing successful and robust AI solutions.