What we build

Custom AI and software for your product or operations, from a first prototype to a system your team runs.

How a project runsDescribe, plan, build, hand over

Four steps, a written plan and a fixed price within days.

Subscriptions

Ready-made workspaces by monthly subscription. Try free, cancel any time from your account.

See plans

No account firstPay, and you are in

Your account is created when you subscribe.

Open research

Reports, guides and methods on how AI can fail, free to cite.

All publications

Latest · 10 Oct 2026Coding agents that deleted production data

Two reported cases, in July 2025 and April 2026, in which an AI coding agent destroyed live data, and what they teach about permissions and backups.

Live AI risk

Automated probes, outage data and public reports, reviewed by a person.

Open the monitor

AI risk nowLevel 1 · Calm

6 systems, 2 regions, 0 reports in 24 h.

Before it fails

What to do in the first 72 hours if AI-dependent services fail.

Read the guides

Offline kitSave it now

Guides and data to keep when the internet is down.

TOSWO

Contact

How to reach TOSWO researchers: press, research, corrections, security and AI behavior reports.

We read everything and answer what we can. Pick the line that fits, it gets you to the right person faster.

You want to	Do this
Report AI behavior you have seen	Use the report form. It can be anonymous. You get a tracking code to check progress.
Ask a press or research question	Write to [email protected]. Say your deadline and we will tell you if we can meet it.
Report an error on this site	Write to [email protected] with the page address and what is wrong. Accepted fixes are listed on the corrections page.
Report a security problem in this site	Follow the security page.
Use our data or logo	See use the data, how to cite us and brand and logo. No permission is needed for the data; the logo has rules.

What to expect

We do not run a help desk. Replies usually take a few working days, longer for requests that need research. We never share the identity of someone who reported AI behavior to us, and we will not forward a private message without asking.