Last updated September 22, 2026
AI Usage Policy
Responsible creative use
Use Dihaw to improve photographic presentation and explore design ideas responsibly. This policy covers uploads, prompts, reference images, DihawGPT conversations, generated media, and public Explore sharing. It applies alongside the Terms & Conditions and provider safety requirements.
Ownership and authorization
Only upload or edit material you own or are licensed to use. Client work requires appropriate permission for AI processing, not just permission to possess the photograph. Respect copyright, trademarks, confidentiality, and publicity rights. Publicly accessible media is not automatically licensed for editing or reuse.
Truthful property marketing
Do not conceal structural damage, damp, unsafe wiring, access restrictions, or other material defects to mislead buyers or tenants. Do not add views, facilities, room sizes, boundaries, or neighboring features and imply they exist. Check local listing-platform requirements before publishing any altered image.
Disclosure and original comparisons
Clearly identify virtual staging, renovation concepts, changed landscaping, and other material edits when sharing. Retain originals for comparison. A prettier result is not proof of current conditions. Label daylight, twilight, sky replacement, and other presentation changes when their omission could mislead viewers.
Portraits and personal privacy
Get appropriate permission before retouching an identifiable person or removing their background. Do not create non-consensual intimate or sexual imagery, impersonate someone deceptively, or produce humiliating, exploitative, or harassing content. Avoid private documents, addresses tied to vulnerable people, access codes, and personal records in reference images.
Children and prohibited harmful content
Never upload, request, or distribute child sexual exploitation material, grooming content, or sexualized depictions of minors. Do not use the service to facilitate violence, criminal activity, targeted threats, hateful abuse, or non-consensual exploitation. Requests may be refused or restricted under provider or platform safeguards.
No deceptive evidence or credentials
Do not fabricate inspections, insurance evidence, identity documents, certifications, permits, or photographs intended to deceive in a legal or financial process. Do not present AI-generated content as verified evidence, an official assessment, or a real person's authentic statement when it is not.
Professional and technical limits
Floor plans, renovation concepts, room layouts, and assistant responses are not construction documents. Have qualified professionals check dimensions, load-bearing changes, accessibility, permits, safety, costs, and feasibility. Enhancement and upscaling cannot guarantee recovery of facts missing from the original image.
Tool-specific expectations
Supply appropriate source material: flambient needs ambient and flash exposures; panorama enhancement needs an existing 2:1 panorama. These tools do not guarantee accurate exposure merging, unseen geometry, a 3D reconstruction, or a virtual tour. Review transparent edges, sky boundaries, reflections, faces, and perspective before delivery.
Explore and third-party reuse
Publishing is opt-in. Confirm you have permission for public viewing and download, including client approval where necessary. Do not publish private or sensitive information. A result's presence in Explore is not an endorsement of its accuracy or a blanket commercial reuse license. Unpublishing does not remove copies already saved elsewhere.
Safeguards and account restrictions
Do not evade content filters, disguise prohibited requests, misuse accounts, or attempt to extract private information from the assistant or other users. A provider rejection may prevent completion. Failed jobs release reserved credits according to the billing rules, including expiry of held monthly credits from an ended cycle. Repeated abuse may lead to restrictions.
Report a concern or request review
Email [email protected] with the result URL or job ID, a short explanation, and relevant rights or safety concerns. Avoid sending unnecessary sensitive material. You may ask for review of an incorrect restriction or publication. We may remove public access or limit processing while investigating; we do not promise that every submission is manually reviewed before publication.