Graham Beale's AI shift research
Privacy notice
Who is responsible for your data
This is a personal research project run by me, Graham Beale. I publish the project under the name Turbulent Ground, but Turbulent Ground is not a company or a separate legal entity.
I'm Graham Beale, and I'm the data controller for this research, the person legally responsible for how your information is collected and used.
ICO registration reference: ZC232927. You can check the public register at ico.org.uk/register.
For anything about this research or your information, email graham@turbulentground.com.
What this is
This research study is for people working in or alongside product development teams who are experiencing the AI shift first-hand. People may take part after a direct invitation or through the public study page. Taking part requires confirming you are 18 or over, checked on the entry screen before the survey starts.
The aim is to understand how AI is changing people's work, and the personal and organisational conditions that help them use it with judgement, agency and care.
I may have your name and email because I know you through my professional network, because someone introduced or recommended you, because you provided them after receiving an invitation, or because you entered them when starting through the public study page.
What I'm asking you to do
- Optionally describe your role, professional discipline, product responsibility and organisation size
- Answer 24 statements about your own contribution and the conditions around you
- Optionally add context to selected answers and a comment in your own words
- Optionally receive an emailed comparison of your answers with the current study benchmark
- Optionally receive occasional study emails, no more than once every 30 days
What I collect, and why
| What | Why | Lawful basis |
|---|---|---|
| Your name, email address, unique invitation token, invitation date and participation status | To select and invite participants, control access, prevent duplicate responses and administer the study | Legitimate interests |
| Your email address and email preferences | To send your requested comparison and, if you choose, occasional follow-up questions and findings | Consent |
| Optional feedback you give when unsubscribing from study emails | To understand why people leave and improve study communications | Legitimate interests |
| Feedback you choose to give about the research study or participant experience, and any contact details you provide with it | To improve the research design, questions, communications and participant experience, and to respond where appropriate | Legitimate interests |
| Your answers to the 24 statements | To conduct the research, create your comparison and calculate anonymised, aggregated study benchmarks and findings | Consent |
| Your optional role, professional discipline, team responsibility and organisation size | To understand whether different kinds of work experience the AI shift differently | Consent |
| Anything you write as optional context for an answer or in the final comment | To understand what sits behind an answer and identify experiences that the fixed statements may not capture | Consent |
| When you started and completed, how many paired themes you answered, and whether your response meets the benchmark inclusion threshold | To administer the study, calculate the benchmark fairly and understand whether the survey is asking too much | Consent and legitimate interests |
| The kind of screen you used (phone, tablet or computer, from the width of your browser window, not your exact screen size, device model or IP address) and the layout version of the survey you saw, at the start and at the end | To check that the survey reads the same way on every device, and to see whether a change to its layout affects answers | Legitimate interests |
| Basic website and server information, such as page visits, request times, device or browser information and security logs | To operate, secure and improve the website | Legitimate interests |
You do not have to take part. The profile questions, answer context and open comment are optional, and each statement also offers “Prefer not to say” and “Not applicable”.
Where I rely on legitimate interests, those interests are administering the study, keeping the survey secure, preventing duplicate submissions and maintaining a reliable website. I limit this processing to what is reasonably necessary and give you the right to object below.
Please do not include information about your physical or mental health, alleged wrongdoing, trade-union membership or any other sensitive personal information. Do not identify a colleague or anyone else by name. If you include this unexpectedly, I will remove it from the research dataset where reasonably possible and will not intentionally use it in the analysis.
Consent, and what each part means
Before the survey, taking part and allowing your responses to contribute to anonymised, aggregated findings are presented as one required agreement because creating those findings is the purpose of the study. A separate, optional checkbox on the same screen lets you choose whether to receive occasional study emails. It is not selected in advance and is not required to complete the study. When requesting your comparison after the survey, a second optional checkbox lets you review or change that email preference. You can also separately say whether you are open to being quoted by name.
- Taking part. Answering the survey and allowing your responses to be analysed and contribute to anonymised, aggregated research findings and benchmarks.
- Receiving occasional study emails. These may include findings, follow-up questions and opportunities to contribute to the research community. I will email no more than once every 30 days. You can unsubscribe at any time and still take part now.
- Receiving your comparison. After completing the study, you can separately ask me to email your answers and, when enough eligible responses exist, compare them with the current study benchmark.
- Being quoted by name. I will never do this without asking you separately, after the fact, about that specific quote.
Saying no to a comparison, study emails or being quoted by name does not affect your participation. Your responses still count.
You can withdraw consent by emailing me. I will stop using and delete or anonymise the personal data covered by that consent. Information already irreversibly anonymised, combined into aggregate statistics or published without identifying you is no longer personal data and cannot be separated back out.
Your personal comparison and the benchmark
If you request it, your comparison shows your answers across 12 themes. It keeps “your contribution” and “conditions around you” separate and may show how each compares with the mean of eligible completed responses.
The benchmark is only shown once there are enough eligible completed responses. It is likely to fluctuate frequently during the early phase of the research. The comparison is descriptive, not a psychological assessment, and it is not used to make decisions about your employment or access to anything.
How your response is linked to you
Your invitation link contains a unique token. The same token links your identity record to your response, controls access to the survey, prevents duplicate submissions and lets me send a requested comparison. Your response is therefore pseudonymous, not anonymous, while I retain that link.
I store identity and contact details separately from survey responses in Airtable. Access is limited to me and the service providers needed to run the study. When analysing response content, I work from the response dataset without names or email addresses wherever reasonably possible.
If you're referred, or you refer someone else
After you complete the study, you can optionally share a personal referral link so others can take part. This is entirely optional and is never required to finish the survey or to receive your comparison. I only issue you a referral link once your own response is complete and eligible.
Your personal link contains a random referral code, not your name, email, answers or identity. Sharing it does not let anyone see anything about you, and starting the survey through someone else's link does not let them see anything about you either — a referral code itself never grants access to anyone's saved answers, submission or identity record.
If you start through someone else's referral link, I keep a record connecting your entry to theirs, so I can understand which recruitment channels and referrals bring in eligible responses. That record links two entries, not two people's answers. If you later start again through a different referral link, I record that as a separate event without changing who was originally credited with referring you.
Where your data goes, including AI
I may use AI tools, including services from OpenAI and Anthropic, to help review the research design, organise pseudonymous responses and identify themes in comments. I do not intentionally provide names or email addresses together with response content. AI output is treated as assistance, not as an automated decision about you, and I review the findings myself.
The main services involved are:
- Airtable: stores identity records, consent records and survey responses in separate tables
- Resend: sends secure survey return links, requested comparisons and study emails
- Vercel: hosts the website and server functions
- OpenAI and Anthropic: may assist with pseudonymous research analysis and review
Some suppliers process data in the United States or other countries outside the UK. Where UK data protection law requires it, restricted transfers are covered by an adequacy regulation or contractual safeguards such as the UK International Data Transfer Addendum or International Data Transfer Agreement. Email me if you would like more information about the safeguards relevant to your data.
How long I keep it
If you start through the public study page but do not finish, I keep the incomplete answers and connected name and email for up to 14 days. I send one reminder after seven days. If you still do not complete the survey, I automatically delete that incomplete response and its connected contact details after 14 days.
I keep identifiable invitation, contact, consent and response records for up to 24 months after you complete the study, unless you withdraw sooner. After that, I delete them or irreversibly remove the link to you. I may keep anonymised aggregate statistics and published findings indefinitely because they no longer identify you.
If you unsubscribe or ask not to be invited again, I may retain the minimum information needed to respect that request. Website and supplier logs are kept for the shorter operational periods set by those services.
Your rights, and how to use them
- Access: ask for a copy of the personal information I hold about you.
- Correction: ask me to correct inaccurate or incomplete information.
- Erasure or restriction: ask me to delete your personal information or limit how I use it, where the law gives you that right.
- Portability: ask for information you provided in a commonly used electronic format, where this right applies.
- Withdraw consent: withdraw consent for research use or study emails at any time, without affecting processing already carried out lawfully.
Your right to object: where I rely on legitimate interests, you can object to that use of your personal information. You can always object to direct marketing. Email graham@turbulentground.com.
To unsubscribe from study emails but keep your existing response in the research, use the unsubscribe link in any study email. The link opens a confirmation page where you can optionally explain why you are leaving. Your preference changes only when you select “Confirm unsubscribe”. This feedback is optional and is used only to improve study communications. You can also contact me with the subject “Unsubscribe from study emails”. To ask me not to invite you again, contact me and I will keep only the minimum suppression record needed to respect that choice.
Questions or concerns
Email me at graham@turbulentground.com. If you are not satisfied with my response, you can complain to the Information Commissioner's Office.
Last updated: 24 October 2026