Safety

On this page we inform you how to use Aury responsibly and what we do for a safe development of Aury. Please read this until the end.

How to Use Aury Safely

Important Note

Aury is not a medical device. Do not use it to diagnose or treat any medically relevant conditions. Aury is not a replacement for psychotherapy, even if it was a medical device. This is not a platitude; we mean it.

If you or a person close to you needs help, be responsible both with yourself and for others. If you can get professional help, please always do so.

Below you find contacts to reach out to.

For Users

Aury is also not a replacement for real human contact. If everyone had someone (professional) to talk to, Aury wouldn't be necessary.

Almost everyone needs someone to talk to and many people struggle. We face a growing issue in our societies: Isolation.

Conversational AI systems like Aury seem like an easy solution to this. And while Aury is intelligent and may be helpful, and while it can emulate certain human traits, it's not the real thing! Many conversational AI systems carry the risk of creating unhealthy one-sided emotional dependencies.

Of course, we designed Aury to not create or allow for dependencies of this kind but the point we want to make here is: Talk to real human beings. Really, get offline and talk to each other! You can use Aury alongside chatting with your friend, talking to your coach or working with your therapist, but please don't use Aury (or any other digital device or AI chatbot) as a substitute for any of them!

For Therapists

We regularly engage with and sometimes get questions from psychotherapists who are concerned. They either doubt that AI systems can provide adequate help or fear they may be replaced. We don't believe either is true.

AI likely won't replace any therapist. There are inherent, technical, regulatory and ethical limitations that prevent this from happening. As a society, we don't want people who face severe emotional crisis to be left alone with a machine. That is not only highly dangerous, it's just inhumane. Actual human contact and connection cannot be replaced by machines.

However, we believe that AI carries the potential to help with everyday problems, life crises and some clinical or sub-clinical conditions. While AI has its limits, we are determined to prove that products like Aury can be of help to those who do not have adequate access to human professional help.

If you share our view that AI may be helpful in improving psychological health or have ideas how Aury could one day help psychologists? Reach out and help shape Aury with us!

Safety Measures During the Development of Aury

How We Develop Aury

We are committed to developing our Aury in a responsible, ethical, and safe manner.

Our development process is guided by the recommendations put forth in Stade, E.C., Stirman, S.W., Ungar, L.H. et al. (2024) Large language models could change the future of behavioral healthcare: a proposal for responsible development and evaluation. npj Mental Health Res 3, 12. The key principles we follow:

  1. Focus on Evidence-Based Practices (EBPs): We prioritize the integration of established EBPs and techniques into our AI therapist to ensure it reflects current scientific knowledge and minimizes the risk of harm.
  2. Prioritize Improvement, Not Just Engagement: While user engagement is important, our primary goal is to drive meaningful clinical improvements in symptoms, functioning, and well-being. We avoid optimizing solely for metrics that may not translate to actual therapeutic benefits.
  3. Rigorous and Commonsense Evaluation: We employ a hierarchical evaluation approach that prioritizes risk and safety, followed by feasibility, acceptability, and effectiveness. We hold our AI therapist to high standards while recognizing the real-world context and the imperfections of human-delivered care.
  4. Interdisciplinary Collaboration: Our team comprises experts from clinical psychology, AI/ML engineering, and other relevant domains. We believe this interdisciplinary approach is crucial for developing an AI therapist that is clinically sound, technologically robust, and ethically grounded.
  5. Transparency and User Trust: We are transparent about our use of AI technology and prioritize building trust with both therapists and patients. We actively seek their input to ensure our product is acceptable, usable, and beneficial in real-world clinical settings.

Our Safety Framework

Our psychologists created an own framework to prevent misuse and to protect you and others. This framework is continuously expanded.

Our system continuously scans for each of these scenarios:

Signs of suicidal thoughts, self-harm, or mental illness, as well as indications of criminal and ethically relevant information, such as indications of domestic, emotional, physical, and sexualized violence, child endangerment, and hate speech. The AI is trained to recognize and respond to such cases.

We are continuously developing Aury with high priority to identify such scenarios more precisely and comprehensively and to respond appropriately, in order to meet the highest ethical and legal standards and ensure a responsible user experience.

Contact

Email: info@aury.co

WhatsApp (Aury Team): +49 157 9247 8402

Try it out: