1. Porcha Woodson, wrongfully arrested by Detroit's facial recognition system in 2023, was in what condition at the time of her arrest?
Correct. Porcha Woodson was eight months pregnant when Detroit arrested her on carjacking and robbery charges based on a facial recognition misidentification.
Woodson was eight months pregnant when she was arrested — the case drew particular attention as one of the most egregious examples of the technology's misuse.
2. Post-processing debiasing of a recidivism algorithm — adjusting thresholds separately for different racial groups — was shown to equalize false positive rates across groups. What was the necessary trade-off?
Correct. The fairness impossibility theorem predicts exactly this: satisfying one fairness criterion (equal false positive rates) while base rates differ between groups requires sacrificing another criterion (overall accuracy). Technical interventions cannot make the trade-off disappear; they make it visible and explicit.
Incorrect. Post-processing to equalize false positive rates did reduce overall predictive accuracy — this is not a workaround failure but a mathematical inevitability. When base rates differ between groups, you cannot simultaneously maximize overall accuracy and equalize error rates across groups. The choice between them is a value decision.
3. The Electronic Privacy Information Center (EPIC) filed an FTC complaint about HireVue's AI video interview scoring in 2019. What outcome followed from a combination of this advocacy and external audit pressure?
Correct. HireVue discontinued facial analysis in 2021 following the EPIC complaint, external audit scrutiny, and sustained advocacy — a documented example of external pressure translating bias concerns into actual product change.
HireVue discontinued its facial analysis component in 2021 — a concrete outcome from sustained external audit pressure and advocacy, illustrating that accountability beyond technical testing can produce real changes.
4. Representation bias in training data refers to:
Correct. Representation bias occurs when some groups appear rarely or not at all in training data. The model has few examples to learn from for those groups, and its performance on them is consequently poor — as demonstrated in the Gender Shades study on face recognition.
Incorrect. Representation bias is not about intent or overweighting. It refers to underrepresentation: certain groups appear so rarely in training data that the model learns too little about them, producing worse performance on precisely those groups.
5. Canada's Algorithmic Impact Assessment requirement caused several federal AI projects to be redesigned before deployment. This is significant because it demonstrates:
Correct. This is the core argument for pre-deployment accountability: intervening during design is structurally cheaper, more effective, and less harmful than trying to patch deployed systems that are already affecting real people. Canada's AIA is one of the few documented cases of this actually working as intended.
Incorrect. Canada's AIA demonstrates that mandatory pre-deployment governance can prevent biased systems from going live — intervening when changes are cheapest, before harm occurs, rather than requiring post-deployment patches that may be structurally inadequate.
6. The Facebook/Meta Fair Housing Act settlement (2022) required the company to:
Correct. The HUD settlement required a structural rebuild of the ad-delivery infrastructure for protected categories — housing, employment, and credit. This is notable as a case where regulatory enforcement mandated the "rebuild" verdict rather than allowing the company to choose a cheaper patch.
Incorrect. The settlement required a full structural rebuild of the ad-delivery system for housing, employment, and credit — not a fine, source code release, or blanket targeting ban.
7. Amazon discontinued its AI recruiting tool in 2017 primarily because:
Correct. The model learned from ten years of predominantly male résumé data that male patterns equated to success, penalising anything coded female.
The system penalised female-coded language because its training data was overwhelmingly from men. This is historical bias embedded through label bias.
8. Using "prior arrests" as an operationalisation of "recidivism risk" introduces bias because:
Correct. This is measurement bias — the operationalisation conflates policing activity with criminal behaviour. Communities that receive more police surveillance generate more arrest data regardless of actual crime rates.
Arrests measure both criminal behaviour and policing intensity. In communities that are over-policed, more criminal behaviour is observed and recorded — but this reflects police deployment, not just crime. The operationalisation is flawed.
9. When a researcher discovers a bias gap, Selbst et al. (2019) warn that technical findings often fail to produce organizational change. What is the mechanism they describe?
Correct. Selbst et al. document how the "abstraction" problem strips technical findings of their context as they move up organizational chains — the nuanced story of conflicting fairness criteria and their different human implications becomes a simplified, decontextualized number.
Selbst et al. describe the abstraction problem: as findings travel up organizational hierarchies, the contextual complexity is stripped away, making it harder for decision-makers to understand what the numbers actually mean for real people.
10. In the Obermeyer et al. study, at the same algorithithmically assigned risk score, Black patients had on average how many more active chronic conditions than white patients?
Correct. The 26.3% figure from Obermeyer et al. in Science (2019) made the magnitude of miscalibration concrete — at the same risk score, Black patients were considerably sicker, meaning the system was systematically underestimating their need for care.
The correct figure is 26.3% — at the same risk score, Black patients had 26.3% more active chronic conditions than white patients, demonstrating severe miscalibration by race.
11. Robert Williams was wrongfully arrested in Detroit in January 2020. What was the basis for his arrest?
Correct. Williams was arrested after Detroit's facial recognition system matched him to a grainy surveillance image from a Shinola store — a match that was wrong.
The arrest was based solely on a facial recognition match — no corroborating evidence placed Williams near the store.
12. The NIST 2019 facial recognition evaluation tested how many algorithms?
Correct. NIST's FRVT 2019 evaluated 189 facial recognition algorithms and found the highest false-positive rates for African American and Asian faces.
NIST evaluated 189 algorithms in its 2019 Face Recognition Vendor Test (FRVT).
13. COMPAS, the recidivism-prediction tool at the center of ProPublica's 2016 investigation, remains in use in multiple U.S. states as of 2024 primarily because:
Correct. COMPAS demonstrates that the fix/scrap framework, even when it clearly points toward scrapping, is not self-executing. Organizational inertia, institutional dependency, and the absence of enforcement capacity have allowed the tool to persist despite the ethical analysis pointing clearly toward retirement.
Incorrect. COMPAS's continued use reflects institutional inertia, not ethical or technical justification. The tool fails the Five-Factor Test on multiple dimensions, has not been shown to be fair by neutral analysis, and has not been subject to enforced retirement — despite ongoing criticism.
14. The Optum healthcare algorithm's racial bias arose because it used healthcare cost as a proxy for health need. Which bias type best describes this?
Correct. Cost and need diverge systematically by race when access to care has been historically constrained. The operationalisation was fundamentally flawed — a construct validity failure.
This is measurement bias. The variable being measured (cost) didn't capture the intended concept (need) equally across racial groups, because historical access barriers meant lower cost ≠ lower need for Black patients.
15. Across all four lessons in this module — facial recognition, economic algorithms, healthcare, and content moderation — what is the most consistent pattern?
Correct. Whether in criminal justice, credit, healthcare, or speech — the consistent pattern is that AI systems ingest historical inequality and produce outputs that perpetuate and scale it, with Black and other marginalized communities bearing disproportionate harm.
The consistent pattern across all four domains: AI systems trained on historically unequal data reproduce that inequality at scale, with disproportionate harm to Black and marginalized communities.
16. How many facial images did Clearview AI claim to have in its database by 2023?
Correct. Clearview claimed over 30 billion images by 2023 — scraped without consent from public websites and social media platforms.
Clearview claimed over 30 billion images — all scraped without the consent of individuals photographed.
17. Pre-processing bias interventions are considered highest leverage because:
Correct. Pre-processing intervenes before the model learns from biased data. By cleaning, reweighting, or replacing the data or features at this stage, bias is prevented from being embedded in the model's weights — rather than being patched after the fact in outputs that already encode it.
Incorrect. Pre-processing's advantage is causal: catching problems before they are encoded into model weights. It typically requires more data access and expertise, not less, and its advantage is independent of regulation or fairness constraints.
18. Model Cards (Mitchell et al., 2019) and Datasheets for Datasets (Gebru et al., 2021) contribute to accountability primarily by:
Correct. Model Cards and Datasheets are documentation frameworks, not technical tools. Their accountability function depends on their being required — by procurement contracts, by regulation, or by institutional policy — creating traceable records of what developers knew and disclosed before deployment.
Incorrect. Model Cards and Datasheets are standardized documentation frameworks. Their value is in creating accountability trails — records of what was known and disclosed — not in encryption, audit replacement, or automated detection.
19. The Optum health algorithm published in Science in 2019 — roughly how many Americans did it affect?
Correct. The researchers estimated the algorithm was in use in systems serving over 200 million Americans at the time of publication.
The study estimated the algorithm affected systems serving over 200 million Americans — making it one of the largest documented instances of AI health bias.
20. Which law did EU regulators cite when issuing fines against Clearview AI?
Correct. Multiple EU data protection authorities found Clearview in violation of GDPR and issued significant fines for scraping biometric data without consent.
EU regulators cited GDPR — the General Data Protection Regulation — which requires consent for processing biometric data.