Arrivals: crash log

Hold Short: an AI go/no-go check. Departures: AI projects

Check in your AI project

Why this board exists

More than 80 percent of AI projects fail, by some estimates. That is twice the failure rate of IT projects without AI. Estimates cited by RAND, 2024

42% of companies scrapped most of their AI projects in 2025, up from 17% a year earlier. S&P Global, 2025

95% of companies saw no measurable profit from their generative AI spend. MIT NANDA, 2025, via Fortune (early, self-reported)

Built from RAND, MIT NANDA and 19 other sources.Most fail for reasons already on file. Pilots run a checklist built from past crashes; this one is built from the accident reports: RAND, MIT NANDA and 19 other sources.

Give it a name so your report says what was checked.

What kind of AI?
Stage
You are

Before departure: Clearance

Which places and decisions does this project reach? Coverage is growing. Your results show what we have checked and what still needs research.

Scroll the board for all project details.

Just looking?

Self-reported. Educational. Not an audit or assurance. Register .
Plain wordsLevel Ground

    Share the pre-flight report

    The official report for your sponsor or team, with the full context. This is exactly what gets sent. It opens in your own app; this page sends nothing.

    WhatsApp Text message Email LinkedIn

    Arrivals: crash log

    Plain words on this board

    From Level Ground, plain explanations of the words people are expected to know.

      How this was built

      What broke?

      Sends from your own email app. Your answers are not included.

      Open email
      Simon Mukooza

      Simon Mukooza

      Say - SY-mohn moo-KOH-ZAH
      AI Operations Partner

      I have run logistics operations from the United States, serving customers across all 50 states, and worked through East Africa and the Gulf. Most of what I know about systems, I learned by watching them fail. This board is built the way pilots build checklists: from the accident reports.

      Also built: Level Ground (plain words for work and tech) and Shadow AI Posture.

      Connect on LinkedIn
      DESTINATION BOARD

      Where are you building or operating it, and who will it reach?

      Select the countries where you build or operate the project and where affected people are. Include more than the server location. For the US, choose each relevant state; add a listed city where it applies.

      A destination appearing here is not proof that its legal coverage is complete.
      DECISION TYPES

      Does it decide or advise on any of these?

      Choose every area it touches. Recommendations to a person still count as advice.

      Have you checked which laws, licences and permits apply in these places?

      Your answers are self-reported and appear in your report. Do not enter confidential identifiers. They do not verify our research or replace local professional advice.
      CLEARANCE DELIVERY

      Rules to check before launch
      WHY THESE CHECKS MATTER

      Real cases and emerging protections

      New technology still meets existing privacy, consumer and other laws. These examples show different legal outcomes, not a finding about your project. Face, voice and likeness protections can involve privacy, personality or publicity rights as well as copyright.