Resources

Interactive tools for the book are available now. Further resources — annotated bibliography, links to key documents and organizations, and further reading — will be added at publication.

Key Risk Documents

We have collected key documents that help frame and contextualize the risk of artificial intellilgence.

They are available here: https://treacheroustech.ca/about/documents/

Interactive Tools

These tools have been created with the assistane of generative AI and are meant as thinking tools rather than definitive statements or claims. Please use them with that in mind.

The CASX Framework — Operational definitions and five-level rating scales for the four CASX dimensions (Capability, Autonomy, Scale, Access), with current frontier system benchmarks.

Regulatory Gate-Checker — Comparative models from nuclear, chemical, and aviation regulation applied to AI oversight, plus an interactive gate-checker for classifying hypothetical AI systems under the book’s regulatory framework.

G20 AI Governance Comparison — Filterable comparison of AI governance approaches across G20 member states.

AI Risk Stance Deck – Flip the cards to explore the AI Risk argument map. Who says what in the rhetoric around AI Risk.

What should governments do? – Exploring the CASX framework with an eye to regulation.

Loss of Control: AI Existential Risk–Interactive Bowtie Assessment - A quantitative risk visualisation of pathways to loss of human control over advanced AI. Modelled using Monte Carlo simulation (n=1,000 trials) with triangular probability distributions across all barrier and threat parameters. Click any element in the diagram to explore its parameters and expert justifications.

Tool–Actor Diagnostic – This framework evaluates technologies by degree of agency: whether they merely execute instructions or begin to act as independent actors within human systems.

Published p(doom) estimates, grouped by epistemic position – How AI researchers, industry leaders, and public intellectuals assess the probability of catastrophic or extinction-level outcomes from artificial intelligence. Grouped by underlying reasoning about whether external control mechanisms are sufficient, not by the numbers alone.

Interactive Tools from others

AI Safety Regulations Map - check out your country by clicking

Failure Mode Atlas - Jacob Ortiz’ conceptual map of AI safety concepts, failure modes, modes of “loss of control”

Rogue AI Tracker - keep up with the latest developments

AI Litigation Tracker - who is suing who over what

Bibliography

Full bibliography from the book

Authors’ web sites

William Leiss: https://leiss.ca

Richard Smith: https://www.sfu.ca/~smith

Tyshenko and Leiss 2026 (full paper behind Chapter 9)

Monte Carlo Probabilistic Assessment of Loss of Human Control Over Advanced AI Using Bowtie Analysis

Michael G. Tyshenko 1,* and William Leiss 2

1 Risk Sciences International. 343, 1505 Laperiere Avenue, Ottawa, Canada. K1Z 7T1

2 Professor Emeritus, School of Policy Studies, Queen’s University, 68 University Avenue, Kingston. Canada.K7L 3N9

Abstract A quantitative risk assessment of human loss of control over advanced AI used a Bowtie diagram extended with fault tree-like and event tree-like analysis. Six primary threats were identified (recursive self-improvement, power seeking, deceptive alignment, loss of corrigibility, off-switch subversion, malicious misuse) and six consequences (systemic infrastructure collapse, economic breakdown, resource shortages, non-human value lock-in, human marginalization, global supply cascade failures). Preventive and mitigative barriers were assigned per pathway from expert literature. Input probabilities (threat base rates and barrier failure-on-demand values) were sourced from experts and modeled with triangular uncertainty distributions. A 1,000-iteration Monte Carlo simulation propagated epistemic uncertainty, yielding a median probability of the top event (loss of human control) of 12.8% (90% CI: 11.3%–14.4%), roughly 1 in 8. The distribution is approximately symmetric with slight positive skew, indicating modest tail risk if barrier failures interact. Conditional on the top event, Expected Severity is 1.85 on a 1–10 scale (90% CI: 1.75–1.96), suggesting mitigation is effective in most scenarios. Results align with expert estimates and demonstrate barrier effects; narrow CIs reflect model consistency. Remaining tail risks support precautionary governance, increased alignment research, iterative risk modeling, and investment in international coordination with robust safety measures to reduce the existential risk of AI loss of control. Keywords: artificial intelligence; existential risk; Monte Carlo; risk assessment; fault tree analysis