Harm, Responsibility and the Road Ahead — Assessment

A fabricated recording appears to show a mayor accepting a bribe. After it is debunked, the mayor's supporters dismiss a separately authenticated recording of an actual procurement violation as another possible fake.

Which assessment best explains the harm across both recordings?

• AThe false recording harms reputation, but its correction restores the evidentiary value of other recordings.
• BThe false recording fabricates evidence and creates an excuse to discount independently authenticated evidence.
• CThe authentic recording warrants continued skepticism because earlier fabrication weakens confidence in the same evidence format.
• DThe dispute mainly reflects insufficient disclosure; labeling the first recording would settle the second.

A fake election-office video announces that polling places are closed. Separately, authentic footage of intimidation is dismissed online because viewers have recently encountered convincing synthetic videos.

Which conclusions are supported by these scenarios? Select all that apply.

• A correction to the closure video can help without fully reversing damage to public trust.
• Withholding judgment on authenticated footage is safer while convincing election deepfakes remain in circulation.
• The fake can disrupt participation even if investigators later establish that it was fabricated.
• Synthetic-media awareness can be exploited to cast doubt on authentic evidence.

A newsroom disproves a synthetic video of ballot destruction. A week later, verified footage of an unrelated election violation is dismissed with the claim that convincing videos are now easy to manufacture.

Assertion: Once the synthetic ballot-destruction video is debunked, its contribution to epistemic erosion necessarily ends. Reason: Awareness of convincing fabrications can provide a pretext for dismissing authentic footage.

• ABoth A and R are true and R is the correct explanation of A
• BBoth A and R are true but R is NOT the correct explanation of A
• CA is true but R is false
• DA is false but R is true

In a transit campaign, a synthetic recording falsely depicts a candidate endorsing fare increases, and a genuine budget recording is later dismissed as AI-generated. In a protest dispute, a fabricated assault video is presented as evidence, and authentic footage of police conduct is then dismissed as equally unreliable.

Match each development to its most direct effect on civic evidence.

A council publishes a correction to a fabricated eviction announcement. Residents now also doubt a genuine emergency housing notice, although the council can verify its publication through independent records.

Which communication strategy best addresses both errors without encouraging blanket trust?

• AExplain the fabrication and provide independently checkable evidence for the genuine notice.
• BRecommend disregarding video announcements until synthetic-media detection becomes reliable.
• CEmphasize that the genuine notice looks more natural than the fabricated announcement.
• DAdvise residents to verify notices by checking for the council's standard logo and publication format.

After a fake ministerial resignation video spreads, a real recording of a minister threatening an inspector is called synthetic. An evaluation team wants to measure more than whether people remember the correction.

Which measurements would directly help assess manufactured uncertainty? Select all that apply.

• The average image-quality score assigned to the fake resignation video
• The change in acceptance of the authenticated inspector recording after exposure to the fake
• The proportion of people who recognize the video-generation software as capable of producing realistic political footage
• The frequency of unsupported 'it could be AI' dismissals of corroborated recordings

A false courtroom clip is withdrawn, but audiences subsequently reject verified courtroom footage and describe every recording as potentially fabricated. An analyst must distinguish this broader damage from the original false claim.

The analyst should classify the broader degradation of shared confidence in evidence as ___.

Investigators trace a specific chain: a fabricated admission of election fraud circulated, was exposed as synthetic, became the basis for claims that recordings are untrustworthy, and was then invoked to reject genuine footage. They want to represent that chain rather than imply every deepfake produces the same outcome.

Arrange the events in the causal sequence documented by the investigators.

An employee's public headshot is used in a fabricated intimate image; a student receives a threat to publish a similar forgery unless they pay; and a teenager's ordinary photograph is submitted to an automated nudification tool. None of the targets agreed to these uses.

Which assessments should guide the response? Select all that apply.

• The employee's case is primarily a reputational dispute rather than NCII because the source photograph was public.
• The employee can suffer serious trauma even though the depicted event never happened.
• The payment demand creates an extortion risk independently of whether the forgery convinces everyone.
• The teenager's case warrants heightened protection, including blocking the exploitative transformation.

A company proposes an automated nudification feature using uploaded portraits. Its initial controls are an uploader checkbox claiming adulthood and a footer saying all outputs are synthetic.

Which assessment is most appropriate before launch?

• AThe footer resolves the main harm because viewers are told no real event occurred.
• BThe checkbox transfers responsibility because users explicitly attest to eligibility.
• CRestricting downloads is sufficient because harm depends on public redistribution.
• DThe controls fail to establish depicted-person consent or reliably protect minors.

A support team handles three cases: a worker targeted with a fabricated intimate image, an adult threatened with publication unless they pay, and a minor targeted through a one-click nudification tool. The tool operator also wants one preventive change that addresses the third case at its source.

Match each response focus to the most directly relevant action.

A studio proposes a fully synthetic character explicitly portrayed as 16 in exploitative sexualized material. Its review board applies industry standards that prohibit depicting minors in simulated sexual activity or nudity.

Assertion: The review board should reject the proposed depiction under those standards. Reason: Those standards protect against prohibited depictions of minors even when no real child's photograph was used.

• ABoth A and R are true and R is the correct explanation of A
• BBoth A and R are true but R is NOT the correct explanation of A
• CA is true but R is false
• DA is false but R is true

A photo application is considering an automated clothing-removal feature for international release. The product manager argues that easy availability of similar tools demonstrates that the feature is acceptable.

Which responses appropriately challenge that launch rationale? Select all that apply.

• Evaluate likeness abuse and block prohibited minor-related transformations regardless of competitor behavior.
• Treat widespread availability as evidence that uploader consent is sufficient.
• Track regulatory action across relevant jurisdictions rather than assuming one country's approach applies worldwide.
• Use notice-and-removal as the primary safeguard until a regulator links the feature to a specific injury.

An attacker sends an adult a fabricated intimate image and demands payment to prevent its publication. A triage analyst initially records only a misleading-media complaint.

The payment demand requires an additional risk classification of ___, even if the image is demonstrably fabricated.

A teenager reports that classmates used an easy-access tool to create an exploitative forgery of them. The moderation team can restrict access immediately but has not yet established the creator's identity.

Which initial response best protects the teenager?

• ARequire proof of the creator's identity before restricting the reported material.
• BAsk the teenager to post a brief denial without reproducing the image while staff verify the report.
• CRestrict access, use confidential child-safety escalation, and avoid unnecessary recirculation.
• DLeave the material available with a synthetic label while staff determine intent.

A worker reports workplace humiliation from an intimate forgery, an adult reports a threat to distribute one to their family, and a minor reports repeated regeneration after takedowns. A manager proposes using view counts as the sole measure of seriousness.

Which observations undermine that proposal? Select all that apply.

• A low-view forgery can still cause serious trauma within a small workplace.
• A threat can coerce its target before the material receives any public views.
• Repeated generation involving a minor signals a prevention failure beyond removal counts.
• A clearly labeled forgery can be assigned low severity when its public view count remains small.

For this educational exercise, LoopShare primarily hosts user-uploaded videos, while EditDesk provides private editing software and a library of company-authored tutorials without user hosting. Apply the stated FTC guidance rather than making a comprehensive legal determination.

Which preliminary Section 3 assessment is best supported?

• ABoth services fit because the hosting criterion includes company-authored media as well as user uploads.
• BEditDesk fits because it enables depiction creation; LoopShare falls under distribution rules rather than Section 3.
• CNeither service is covered unless it created the reported depiction.
• DLoopShare fits the stated hosting criterion; EditDesk does not on these facts.

LoopShare is treated as a covered platform in this educational scenario. At 10:00 a.m. Monday it receives a valid, signed request identifying a covered depiction, and it already knows of two identical copies on its service.

Which actions follow from the stated Section 3 summary? Select all that apply.

• Remove the identified depiction within 48 hours of receiving the request.
• Start the 48-hour period when a moderator first opens the ticket.
• Remove the known identical copies within the same 48-hour period.
• Treat removal of the requested URL alone as sufficient despite the known copies.

LoopShare mainly hosts user posts; EditDesk offers private editing and company-authored tutorials without user hosting. An educational compliance note also identifies a valid signed request received Monday at noon and the federal enforcer of Section 3.

Match each item to the conclusion supported by the stated summary.

In an educational compliance exercise, a covered platform receives a valid, signed request at 3:00 p.m. Friday. Its support team normally works Monday through Friday, and it knows of an identical copy in another user's account.

Which response meets the timing and scope described in the summary?

• ARemove the original by Tuesday at 3:00 p.m. and assess the copy afterward.
• BRemove the depiction and known identical copy by Sunday at 3:00 p.m.
• CRemove the original by Sunday and start a separate clock for the copy.
• DRemove both within 48 hours after the Monday support shift begins.

An educational brief assigns Section 3 enforcement to the FTC. It separately explains that convincing synthetic media can cause severe personal harm.

Assertion: The FTC is the enforcer of the Section 3 platform duties described in this brief. Reason: Convincing synthetic media can cause severe personal harm.

• ABoth A and R are true and R is the correct explanation of A
• BBoth A and R are true but R is NOT the correct explanation of A
• CA is true but R is false
• DA is false but R is true

A covered platform receives a valid, signed request at noon Monday but assigns it to a reviewer at noon Tuesday. An engineer is implementing the compliance timer using the educational summary.

The removal deadline must be calculated from request ___, rather than assignment to a reviewer.

LoopShare primarily hosts user content, while EditDesk only supplies private editing and company-authored tutorials. A product team is turning the educational TAKE IT DOWN Act summary into an internal note, not legal advice.

Which statements belong in that note? Select all that apply.

• FTC enforcement of Section 3 concerns the platform duties described in the summary.
• EditDesk can defer abuse-prevention controls to hosting platforms because it falls outside the stated hosting criterion.
• LoopShare's response must include known identical copies when the stated valid-request requirement applies.
• Borderline service designs require further assessment rather than certainty from this simplified comparison.

A covered platform is implementing a serial audit trail for an already-valid, signed request identifying covered material. Its design requires a receipt record before a removal job, and a removal-result record before ticket closure; the removal deadline remains 48 hours from receipt.

Order the records and actions in this specified audit trail.

A studio is creating a synthetic spokesperson modeled on a paid adult performer. It plans cryptographic watermarking, a consent form, and a synthetic-media label, and needs release checks that assess the safeguards separately.

Which checks provide meaningful evidence that the safeguards were implemented? Select all that apply.

• Accept the verified studio watermark as confirmation that the linked performer release covers the final script.
• Confirm that written permission covers the planned likeness use, distribution, and material transformations.
• Validate the cryptographic provenance signal on the actual distributed files.
• Inspect disclosure visibility in the player and on supported resharing surfaces.

A performer signed permission for a synthetic product demonstration. The marketing team now wants to make the avatar deliver a political endorsement, while retaining the studio's authenticated watermark and synthetic-media label.

Which release decision best respects the three safeguards?

• ARelease because authentication links the new video to the same authorized production.
• BRelease because disclosure prevents viewers from mistaking the endorsement for a live appearance.
• CPause and obtain meaningful written permission covering the new use before proceeding.
• DRelease after checking that the political script contains no demonstrably false claims.

A campaign's cryptographic watermark validates successfully, but its synthetic spokesperson makes an unsupported medical claim. The reviewer must state exactly what the successful check establishes.

The watermark check supports ___, not the medical claim's truth or the performer's permission.

A studio's synthetic training presenter will appear in a web player and downloadable clips. The workflow includes cryptographic watermarking, written likeness permission, visible synthetic-media disclosure, and a separate review of factual statements.

Match each implementation step to its most appropriate verification method.

A synthetic spokesperson video has valid written permission and a cryptographically verified master file. In mobile previews, the synthetic-media label is cropped out, and the distributed transcode has not been checked for watermark preservation.

Which changes directly address the unresolved release risks? Select all that apply.

• Rely on the consent form because it also authorizes the studio's disclosure choices.
• Add a press-release statement instead of changing the cropped viewing experience.
• Validate provenance on the distributed transcode rather than relying only on the master.
• Reposition or supplement disclosure and verify its visibility in the mobile preview.

A studio adds a valid cryptographic watermark to a synthetic interview. Its release checklist says that watermark validation also proves every interview statement is true.

Assertion: The studio should keep the watermark as a provenance safeguard. Reason: A valid cryptographic watermark establishes that every statement in the interview is true.

• ABoth A and R are true and R is the correct explanation of A
• BBoth A and R are true but R is NOT the correct explanation of A
• CA is true but R is false
• DA is false but R is true

A synthetic presenter passed provenance and disclosure checks, but the signed agreement authorized internal training only. The proposed release is a public advertisement.

Before release, the team must obtain appropriately scoped written ___ rather than infer authorization from the technical checks.

A studio is choosing a release package for a synthetic product demonstrator. It needs implementation evidence and verification evidence for provenance, consent, and disclosure, without allowing one safeguard to stand in for another.

Which package best meets that requirement?

• AWatermark and validate the export; obtain scoped written consent and compare it with final use; add disclosure and inspect published views.
• BWatermark and validate the export; infer consent from the performer's upload; add disclosure and inspect the published views.
• CWatermark and validate the master; archive scoped written consent; add disclosure and approve its placement using the editing preview.
• DWatermark the export; obtain scoped written consent; add disclosure; verify all three by confirming the publisher accepted the completed release checklist.

A generator has verified written consent and cryptographic provenance, but its clips lack synthetic labels, its minor-safety tests are absent, complaints require a paid account, and only every hundredth output gets human review. The audit must address all four remaining gaps.

Match each gap to a corrective action that directly addresses it.

A studio uses a visible logo as its only provenance measure and accepts an uploader's 'rights cleared' checkbox instead of written permission from depicted people. It already has tested disclosure, minor-protection controls, accessible reporting, and human oversight covering all outputs.

Which corrections address the two remaining checklist gaps? Select all that apply.

• Embed a unique studio logo in every export and keep a matching publication log for provenance verification.
• Implement cryptographic watermarking and verify it on distributed outputs.
• Obtain meaningful written consent and verify its scope against each intended use.
• Retain the uploader's rights declaration as written consent and audit a sample of declarations after release.

A service documents provenance, written consent, disclosure, and protection of minors. Its remaining controls are a complaint form that requires a creator account and a safety classifier whose decisions no employee is assigned to review.

Which change best closes both remaining gaps?

• AProvide simplified creator signup for reporters and raise the classifier's confidence threshold for automatic release.
• BPublish classifier accuracy, add reporting instructions to signup, and assign an owner to the annual policy review.
• CLet creators forward complaints privately and review a random monthly sample.
• DEnable reporting without a creator account and assign human reviewers with escalation authority.

A service records all generated outputs and runs automated safety checks, but nobody has authority to review decisions, pause release, or escalate harm reports. Its other five creator safeguards are implemented.

The unresolved safeguard is continuous ___, not additional automated logging alone.

A studio's workflow has cryptographic provenance, scoped written consent, and appropriate disclosure. It lacks minor-protection testing, requires a government ID for every initial harm report, and leaves low-risk classifier outputs outside any human oversight process.

Which audit recommendations address the missing safeguards? Select all that apply.

• Add minor-protection controls and test prohibited requests, including attempts to bypass them.
• Offer an accessible initial reporting route with proportionate, privacy-conscious follow-up.
• Extend accountable human oversight and escalation procedures to all output paths.
• Use signed uploader assurances for minor protection and retain identity-verified reporting to deter fraudulent complaints.

An audit finds functioning minor protections and human oversight, but identifies four defects: watermark validation fails after export, permission covers only internal use, public previews hide the synthetic label, and the report link returns an error.

Match each defect to the verification evidence needed to close it.

A high-volume service satisfies provenance, consent, disclosure, minors protection, and reporting requirements. Its automated generation pipeline has no named human owner for most outputs, and reports about those outputs remain in an unmonitored queue.

Which redesign best meets the remaining oversight requirement?

• AAssign a safety lead to approve each model version and treat that approval as oversight for all outputs until the next update.
• BCover every output path with accountable human oversight, risk-based release review, and monitored escalation with stop authority.
• CAssign a legal-response owner to review challenged outputs after formal notices arrive, using automated screening for all releases in the meantime.
• DAssign reviewers to high-reach outputs and monitor aggregate classifier performance, keeping remaining output paths under automated controls without individual human ownership.

Before relaunch, a studio must audit all six safeguards, implement corrections, and complete an independent verification gate. Its procedure requires an evidence inventory before gap mapping and prohibits sign-off until corrected controls pass tests.

Order the steps in this corrective audit procedure.

An intimate forgery of a teacher spreads on a covered platform, which receives a valid, signed request identifying it. The creator used a commercial generator, and known identical copies are already circulating on the platform.

Which division of work best reflects shared responsibility?

• APlatform removes the depiction and known copies within 48 hours and monitors reuploads; regulators enforce and examine systemic failures; creators halt distribution and fix consent and safety controls.
• BPlatform meets the removal deadline and closes the case; regulators handle future prevention; creators wait for a formal finding before changing their workflow.
• CPlatform retains disputed copies pending attribution and preserves evidence; regulators identify the responsible creator and resolve authenticity disputes; creators add disclosure and let the platform coordinate further response.
• DPlatform forwards requests for creator-led cleanup and tracks completion; regulators mediate disputed permissions and issue guidance; creators determine which copies fall within authorized use and submit removal confirmations.

A platform removes a harmful synthetic depiction and known identical copies 30 hours after a valid, signed request. The target continues receiving threats, and the creator's workflow still allows the same unauthorized likeness use.

Which follow-up conclusions are justified? Select all that apply.

• Timely removal supports closing the platform's incident review while routing continuing threats and workflow concerns to the creator.
• The platform should keep an accessible reporting route and escalate continuing threats to trained humans.
• The creator should review how permission and misuse controls failed rather than treating removal as remediation of the workflow.
• Regulators can examine compliance patterns and systemic risks without taking over the creator's day-to-day release decisions.

After a synthetic-media abuse incident, the generator operator says that the hosting platform's removal process makes further production safeguards unnecessary. The platform separately argues that the creator's misconduct excuses delayed removal.

Assertion: Shared responsibility does not permit either actor to outsource its own responsibilities to the other. Reason: Upstream prevention and downstream removal address different stages of harm and can both be necessary.

• ABoth A and R are true and R is the correct explanation of A
• BBoth A and R are true but R is NOT the correct explanation of A
• CA is true but R is false
• DA is false but R is true

A synthetic impersonation incident requires six work packages: urgent platform removal, continuing platform reporting, regulatory enforcement, regulatory prevention work, creator containment, and creator workflow repair. The platform's valid-request duty is treated separately from broader ethical measures.

Match each work package to its actor and concrete actions.

An incident team wants to publish a public explanation of an NCII forgery, retain evidence for authorized review, and reassure the affected person. Some staff propose embedding the image in the public post so readers can verify what was removed.

Which actions better protect human dignity while preserving accountability? Select all that apply.

• Keep necessary evidence in access-controlled records rather than making the target's image a public exhibit.
• Publish the labeled forgery in a click-to-view section of the public report so readers can independently verify the incident.
• Ask the affected person to confirm their account publicly before releasing an update, so the response does not appear unsubstantiated.
• Provide a non-graphic explanation of actions taken and a private route for the affected person to raise continuing concerns.

A platform satisfies the stated valid-request removal duty, then voluntarily supports the target, investigates repeat abuse, and improves prevention with creators. A manager incorrectly describes all these actions as requirements of the same 48-hour rule.

The plan should distinguish the legal removal duty from broader ___ responsibilities.

After a likeness-abuse incident, a joint review must choose measures that show whether its response plan works. Platforms own removal and reporting, regulators own enforcement review, and creators own production controls and human release oversight.

Which measures would provide useful evidence across these responsibilities? Select all that apply.

• Time from valid-request receipt to removal of the depiction and known identical copies
• Number of synthetic labels linked to a creator's rights declaration, used as proof that likeness permission was obtained
• Repeat-abuse findings and documented follow-through from regulatory compliance reviews
• Results of creator consent-scope checks and records of human intervention in unsafe releases

A shared-response plan already assigns platforms timely removal and accessible reporting, regulators compliance review and systemic investigations, and creators distribution pauses and production-control repairs. The team must decide how to govern unresolved threats and future failures after the first incident closes.

Which addition best makes the plan durable?

• AMove follow-up into routine compliance reporting after timely removal, reopening joint review only when a new formal complaint establishes a breach.
• BGive regulators ownership of recurrence monitoring and have platforms and creators act on referred findings rather than maintain parallel monitoring processes.
• CAssign ongoing human owners, review recurrence and target feedback, and verify corrective controls across the three roles.
• DPublish a joint transparency statement with corrective-action commitments and use completion of that publication as the final cross-role accountability check.