Tools.
AI safety messenger map.
A directory of people and organisations explaining AI risk to the public, across both the existential-risk and present-day-harms conversations. Searchable and filterable by focus, type, region and channel.
AI attitudes segment calculator.
A working demo of a segmentation calculator: answer ten questions and see where you land. The segments are illustrative placeholders, not the output of our pilot study.
Concept demoAI Comms Toolkit.
A practical toolkit, built from tested evidence. Free, open guidance that any communicator can pick up and use.
In developmentBuilt from the research.
The toolkit is the first thing the research is for. It comes directly out of the pilot's findings, so nothing in it is published until it's been tested.
- 01Tested message guides, broken down by audience segment
- 02Do's and don'ts for common claims about AI, grounded in what the data actually shows
- 03Plain‑English summaries of the research, for people who need the answer
Anyone talking about AI who wants to stop guessing.
Campaigners and communicators across both the present‑day‑harms and existential‑risk conversations, policy teams drafting public‑facing material, journalists, and researchers who need their findings translated into something usable.
Frequently asked questions
What is in the toolkit?
Tested message guides broken down by audience segment, a messenger map showing who each segment trusts and how to reach them, guidance on which claims about AI to make and which to avoid, and plain-English summaries of the research. All of it will be free and open.
Is it available now?
Not yet. The toolkit is built from the pilot findings, so version 1 follows the pilot. We will not publish message guides for messages we have not tested. You can register for early access by emailing [email protected].
Will it cost anything?
No. The toolkit is free and stays free. That rule does not change as Common Signals grows.
Who is it for?
Anyone talking about AI who wants to stop guessing: campaigners and communicators on both the present-day-harms and existential-risk sides, policy teams, journalists and researchers who need findings translated into something a comms team can pick up and use.
How is this different from advice I could get from a comms agency?
It is built on published evidence rather than instinct, it is open rather than proprietary, and it is neutral: the guides tell you what worked with which audience, not what position to take. An agency can and should use it too.
Will the toolkit be updated?
Yes. Each new study adds to it, and the plan includes a live public-attitudes tracker so the guidance keeps pace as the technology and the public conversation change.
Comment analyser.
Paste a Hacker News link or upload a comment export from any platform, and get a coded read of how a post landed: stance, frames, what carried the engagement, and the charts to show it. The same method as our published analyses, run on your thread.
In development