NIST AI RMF, ManageMANAGE 4.1

On receiving a valid removal request, remove the image and make reasonable efforts to remove copies within forty-eight hours.

Remove a validly requested altered sexual depiction, and make reasonable efforts to remove known identical copies, within 48 hours of receiving the request.

As a telecommunications or electronic communications network operator or service provider, act to bar public access to such an image once you become aware of it through your professional activity, and make it available to the judicial authority.

If you own a website, application, or social media platform on which such material is disclosed, you are liable for damages if the depicted person requests removal and you fail to remove it, and known identical copies, within 72 hours.

Give users a clear way to report a non-consensual counterfeit intimate image and remove a reported image within 48 hours of notice, if you operate a covered platform

An operator of a website, online service, or application may not accept a fee or other consideration to remove, delete, or refrain from posting such an image when the depicted person requests it.

Let a consumer opt out of your use of ADMT to make a significant decision about them, unless you offer an appeal to a human reviewer with authority to overturn the decision or another exception listed in the regulation applies

Provide the system so the assigned overseers can decide, in a particular situation, not to use it, or can disregard, override or reverse its output.

If you are a financial institution subject to internal-governance requirements under Union financial services law, you meet this duty by complying with those internal-governance rules, except that you must still have the risk management system, the post-market monitoring system and the serious-incident reporting procedures of Article 17(1), points (g), (h) and (i); harmonised standards referred to in Article 40 are taken into account to that end.

Undergo a new conformity assessment procedure if you substantially modify a high-risk AI system that has already been assessed, whether or not the modified system is intended to be further distributed or continues to be used by the current deployer; for a system that continues to learn after it is placed on the market or put into service, changes you pre-determined at the initial assessment and set out in the technical documentation under point 2(f) of Annex IV are not a substantial modification.

If you are the provider of a high-risk AI system, establish and document a post-market monitoring system proportionate to the nature of the AI technologies and the risks of the system.

Base the post-market monitoring system on a post-market monitoring plan, and make the plan part of the technical documentation referred to in Annex IV.

Continuously keep your AI system safe, secure, and reliable, and promptly detect and remedy any incident capable of harming people, property, data, or social order, regardless of the system's risk tier.

If you discover a user has used the service to engage in illegal activity, warn the user, restrict their access to features, or suspend or terminate their access, in accordance with law and your service agreement.

Separately take measures against the responsible user, such as warning, restricting features, suspending the service, or closing the account, in accordance with law and your service agreement.

On receiving a report of explicit deep fake material, immediately confirm to the user that you are aware of it, remove the reported content and any identical known copies, and provide the user a written status update within seven days.

If you determine reported content is explicit deep fake material, take measures to keep the same material from being posted again.

Keep records of any material change made to the software used for algorithmic trading, allowing you to determine when a change was made, the person that made it, the person that approved it and the nature of the change (Article 5(7)); have a person designated by senior management review any proposed material change to the production environment related to algorithmic trading before it is made, and communicate any change to the functionality of your systems to the traders in charge of the trading algorithm and to the compliance and risk management functions (Article 11).

The CBJ expects you to offer opt-out or alternative service options to customers who prefer not to interact with AI-driven systems, without compromising their access to essential banking services, to let customers access, correct and request deletion of their personal data where applicable, and to keep open channels for customers to question and appeal AI decisions that affect them (sections 2.3 and 2.4).

Delete an applicant's interview video, and instruct anyone else who received a copy to do the same, within 30 days of the applicant's request.

Where a violation is found, remove AI-generated programme content as soon as possible and at the latest within one working day of a report (Article 67).

As an internet access provider, remove or block access to illicit content as soon as you become aware of it, directly or indirectly (not only on judicial order), and assist the competent judicial authorities in an investigation.

Remove information as soon as practicable, and notify the Minister or the appropriate law enforcement agency, once you acquire actual knowledge that it gives rise to civil or criminal liability.

If you become aware of facts suggesting a likelihood of liability, follow an approved code of conduct or notify the Minister, who may then direct removal.

Stop processing an individual's already-public personal information once they expressly decline, even though no prior consent was required.