Causal Profile Mismatch
Comparing full causal taxonomy distributions between risks (academic predictions) and incidents (real-world outcomes) per subdomain. Percentages use coded-only denominators, excluding “Not coded” records for cleaner comparison.
How to read this
Each subdomain shows paired stacked bars comparing risk (R) and incident (I) distributions for a selected causal dimension. Use the dimension picker to switch between entity, intent, and timing. Hover over bar segments to see exact percentages and counts. Click any row to see the full breakdown across all dimensions.
Key Takeaways
- 1.1.2 Toxic content has the largest reliable mismatch (Unintentional +53pp in intent).
- 2.Incidents are systematically more AI system than risks predict in 80% of subdomains (median gap: +28.8pp).
- 3.Incidents are systematically more Intentional than risks predict in 87% of subdomains (median gap: +14.2pp).
- 4.9 subdomains with fewer than 5 incidents are confidence-weighted. Percentages use coded-only denominators (excluding "Not coded" records).
Discrimination & Toxicity
1.2 Toxic content
76 coded risks · 87 coded incidents
1.3 Unequal performance
17 coded risks · 33 coded incidents
1.1 Discrimination
82 coded risks · 117 coded incidents
Privacy & Security
2.2 AI security vulnerabilities
111 coded risks · 18 coded incidents
2.1 Loss of privacy
77 coded risks · 84 coded incidents
Misinformation
3.1 False information
53 coded risks · 174 coded incidents
Malicious Actors & Misuse
4.1 Disinformation & influence
82 coded risks · 126 coded incidents
4.3 AI fraud & scams
77 coded risks · 353 coded incidents
4.2 AI weapons & cyberattacks
80 coded risks · 12 coded incidents
Human-Computer Interaction
5.1 Overreliance & unsafe use
60 coded risks · 32 coded incidents
Socioeconomic & Environmental
6.1 Power centralization
51 coded risks · 6 coded incidents
6.2 Inequality & unemployment
54 coded risks · 6 coded incidents
6.3 Devaluation of human creativity
31 coded risks · 5 coded incidents
AI System Safety, Failures & Limitations
7.3 Capability & robustness
123 coded risks · 292 coded incidents
7.4 Transparency & interpretability
41 coded risks · 5 coded incidents