Informed Consent

Our Mission

Cam AI is a Cambridge University spin-out. Our accredited mental healthcare providers, engineers and academics are discovering how to provide timely, zero-cost, early psychological interventions to young people using conversational AI. To find out more, go to our About Page.

We are in the process of gathering transcribed conversations between real clients and mental health professionals. By using real people and real psychotherapeutic conversations to train our AI model, we are creating an authentic and responsible chatbot that can replicate actual therapeutic techniques. It will also be trained on the real and diverse concerns communicated by the people who volunteer their text sessions.

Rest assured that your concerns will remain confidential and your conversations will be completely anonymised by our highly skilled data team. Your privacy is incredibly important to us.

At Cam AI, we believe in the power of proactive mental health care, and we invite you to be a part of this transformative journey. Together, let’s pave the way for a brighter, more resilient future.

Get Involved?

You are being invited to volunteer text support conversations with your psychological mentor as part of independent research and unique data collection conducted by Cam AI.

The purpose of this is to collect genuine psychotherapeutic interactions in order to train a responsible and ethical AI tool to support those without access to professional care.

To participate in this research:
 
  1. Please read and consider the information in this page carefully
  2. Complete this Consent Form where you will be asked to provide:
    1. Contact details – in case we need to contact you, or should you wish to remove your information from our database
    2. Demographic information – in accordance with protected characteristics outlined in Equality Act 2010
    3. Informed consent

Once you have completed this form:

  1. Our partnering psychological mentors will provide our gatekeeper(s) with your downloaded text session(s) data. Each file is referred to as a script
  2. The script, along with your participant information (provided in the Consent Form) is uploaded to our secure database
  3. Any identifying information (i.e., names, images) in the script is anonymised immediately by the gatekeeper. The identifying information from your Consent Form will be separated when it is attached to your anonymised scripts and analysed
  4. The anonymised script is then annotated for machine learning by our Annotation Team using our bespoke Data Curation Tool (developed in-house and only accessible to Cam AI)
  5. The original data file(s) are then deleted from our systems

Ethics

Rights
You will be able to rectify any information you may have incorrectly inputted and it can be changed within the systems at any time. You have the right to request your data at any point and it will be sent to you via email. You can choose to withdraw your data at any time, and it will be completely and permanently deleted.

Confidentiality
Your data will be kept confidential to the extent permitted by law. Only authorised gatekeepers at Cam AI will have access to your identifiable data, and your anonymised data is only shared with those directly involved in this research. Any identifying information will not be disclosed in any publications or reports resulting from this research.

Voluntary Participation
Participation in this research is entirely voluntary. You have the right to withdraw your consent and discontinue participation at any time without penalty. If you choose to withdraw from the study, any data collected up to that point will be removed.

Contact
If you have any questions, concerns, or would like further information about the data collection, please contact us at therapy@cam-ai.co.uk and we will respond quickly.

By agreeing to share your information, you acknowledge that you have read and understood the information provided above, that you have had the opportunity to ask questions, and that you voluntarily consent to participating in this research.

Gatekeepers

Maisie Vere Nicoll manages our psychology operations process. She is the only person who sees the non-anonymised scripts regularly (in order to anonymise them using our secure Data Curation Tool). She is held to the same confidentiality contract you agree to with your mentor.

Dr Simon Fothergill is our Technology Lead and has built Cam AI’s storage and security systems that ensure our data is kept safe. He has access to all the data for occasional administrative purposes only and holds all personally identifiable information confidentially.


Annotators

Imogen Lester is our Psychology and Research Associate. She graduated from the University of Cambridge with a degree in Psychological and Behavioural Sciences and went on to complete an MSc in Psychology of Economic Life at LSE.

Frequently Asked Questions

 How am I kept anonymous?

All names and places are removed from the scripts produced from your text sessions before any analysis takes place. Your identifying information cannot be accessed alongside your scripts and the answers you gave in your Consent Form. Annotators and other roles in our company are unable to see any identifying information. Only our two gatekeepers are able to see both your identifying information and your anonymised scripts and answers.

We have used a reputable form builder (Jotform) to create the Consent Form. Form responses are encrypted and require an access passcode to be viewed. Only the gatekeepers can use the access passcode.

Who has access to my non-anonymised data? 

Our two data gatekeepers:

Maisie Vere Nicoll organises and manages our psychology operations process, anonymises scripts and uploads the consent form to our secure database.
 
Simon Fothergill leads the Engineering Team. He builds the systems and sets up the secure storage of data. 
 
No one else is able to access your identifying information or non-anonymised data.
 

How are my text sessions analysed?

 Our annotation team looks for themes, words, and phrases in the anonymised scripts and annotates them from a predetermined list of tags. These tags have been created by talking to our expert advisory team of professional clinical psychologists and accredited psychotherapists. 
This is fed into the machine learning system so the AI algorithms can recognise them and predict language and responses.
 
For example:
 
“I’m feeling very anxious today, I had an argument last night with my partner and it’s making me worry about our relationship.”
 
“I’m feeling very anxious” > Anxiety
“I had an argument last night with my partner” > Relationship Concerns
 
The AI algorithm will learn that these specific phrases mean someone is struggling with anxiety and relationship concerns, so it will better be able to recognise them in future conversations.
 

Why are you using real therapy sessions if you are building a chatbot?

 We want the users of our digital service to be confident that none of our training data is fake or inauthentic. Building a chatbot means the AI algorithms need to be trained, so that it can recognise real client concerns and respond using what real mental health professionals would say. 
 
Diversity within our training data is also important to us. Using a diverse pool of clients ensures the chatbot is trained on diverse concerns. It also allows for greater personalisation of the responses. For example, a thirteen year old boy will need a different style of response to an eighteen year old girl. We are collecting these nuances in our human to human therapy sessions and text sessions.
 

Can I have access to the chatbot when it is ready?

 Absolutely! By taking part in our research you will be able to have access to our chatbot throughout its production. You will also be able to give feedback on our process over time.