---
title: Photographic Integrity
description: Can AI detectors mistake real photos vs AI-generated images? Explore photographic authenticity, TvF accusations, & how photographers can defend their work.
image: https://flyboyz.net/hubfs/20261101_151525.jpg
---

[Skip to content](https://flyboyz.net/fly-boyz-llc-blog/photographic-integrity#main-content)

![ChatGPT Image Jun 20, 2026, 08\_20\_53 AM](https://flyboyz.net/hs-fs/hubfs/ChatGPT%20Image%20Jun%2020%2c%202026%2c%2008_20_53%20AM.png?width=1553&height=1037&name=ChatGPT%20Image%20Jun%2020%2c%202026%2c%2008_20_53%20AM.png "ChatGPT Image Jun 20, 2026, 08_20_53 AM")

- [Home](https://flyboyz.net)
- Aerial Services 
    - [Precision Aerial Media](https://flyboyz.net/precision-aerial-media)
    - [Mapping & Data](https://flyboyz.net/3d-services)
    - [Cinematography](https://flyboyz.net/drone-cinematography)
    - [Drone Education](https://flyboyz.net/drone-flight-training)
- Business Services 
    - [Marketing & Events](https://flyboyz.net/events)
    - [Architectural Photography](https://flyboyz.net/architectural-photography)
    - [Commercial Photography](https://flyboyz.net/commercial-photography)
    - [Residential Real Estate](https://flyboyz.net/residential-real-estate)
    - [Headshots](https://flyboyz.net/headshots)
- Personal Services 
    - [Weddings and More](https://flyboyz.net/weddings)
- Shop 
    - [Studio Prints](https://santafe.shopwhereilive.com/vendor/fly-boyz-llc/)
    - [NM Skies Collection](https://flyboyz.net/nm-skies-collection)
- Fine Art 
    - [Artist Manifesto](https://flyboyz.net/manifesto)
    - [NM Skies Collection](https://flyboyz.net/nm-skies-collection)
    - [Justina](https://flyboyz.net/justina)
    - [Fine Art Portraiture](https://flyboyz.net/conceptual-fine-art-portraiture)
- About 
    - [Meet the Team](https://flyboyz.net/about)
    - [Portfolio](https://sheabenderportfolio.my.canva.site/)

 Nov 1, 2026, 5:21:21 PM

# Photographic Integrity

[Shea Bender](https://flyboyz.net/fly-boyz-llc-blog/author/shea-bender)

Share: [facebook-f icon](http://www.facebook.com/share.php?u=https://flyboyz.net/fly-boyz-llc-blog/photographic-integrity) [linkedin-in icon](http://www.linkedin.com/shareArticle?mini=true&url=https://flyboyz.net/fly-boyz-llc-blog/photographic-integrity) [Twitter icon](https://twitter.com/intent/tweet?url=https://flyboyz.net/fly-boyz-llc-blog/photographic-integrity) [pinterest-p icon](http://pinterest.com/pin/create/link/?url=https://flyboyz.net/fly-boyz-llc-blog/photographic-integrity) [envelope icon](mailto:?body=https://flyboyz.net/fly-boyz-llc-blog/photographic-integrity)

These days AI is everywhere. It's used in emails, writing, and also in photography. Adobe and other editors have added AI as an option for editing workflows, and as most people know, AIs such as ChatGPT, Google Gemini, and others are becoming increasingly good at generating imagery.

More and more for photographers, the question isn't just whether a photographer can create a convincing image; it's whether they can establish the image's origins when someone challenges its authenticity. This is something that I've experienced myself recently, and looking deeper into it, I found this topic goes much deeper.

In 2023, an Australian photography contest [disqualified photographer Suzi Dougherty's](https://news.artnet.com/art-world/australian-photographer-disqualified-ai-generated-2337906?) photo after being mistakenly deemed to have been generated by AI. Dougherty took her picture as part of several setups on an iPhone, and is not a professional photographer and said in the story that she doesn't "really follow AI".

In another case in 2024, Miles Astray [entered an authentic photo in an AI photo competition](https://www.cbsnews.com/news/real-photo-ai-competition-flamingone-miles-astray/?), where his authentically taken photo (as opposed to an AI generation) won the AI competition. His stated purpose for this was to show that there's a human and emotional quality that AI cannot generate. 

Institutions are increasingly struggling with the consequences of false AI accusations. Among the newer tools being used to address this challenge are AI detectors, which analyze submitted images or written content and estimate the likelihood that they were generated using artificial intelligence. These assessments are probabilistic rather than definitive, and generally evaluate the submitted work rather than reliably identifying every individual AI-generated element within it. [A 2025 peer-reviewed study examined three publicly available AI detectors using authentic and AI-generated scientific images.](https://pubmed.ncbi.nlm.nih.gov/39989748/) This study showed how accuracy in assessment can be varied. It depends on factors such as the detector, the actual content being analyzed, as well as the circumstances in which it's used.  In my own experience, I was assessed based on a [holiday-styled photo](https://drive.google.com/file/d/1X67nzKIhFU3F9RqyHgvbqjahgwTaPpmd/view?usp=sharing) captured with a Sony A7III, which contains a synthetic/AI-generated image on a laptop screen. Several AI detectors returned conflicting, lower-confidence judgments when different images were submitted.

Additionally, several major US universities have explicitly warned against treating AI detector scores as evidence of misconduct. [Caltech](https://aiinteaching.caltech.edu/?) strongly discourages AI-writing detectors because of false positives, false negatives, and damage to trust; [Cornell](https://teaching.cornell.edu/generative-artificial-intelligence/ai-academic-integrity?) considers current detection tech unlikely to provide a workable solution to identifying authorship; the [University of Southern California](https://academicintegrity.usc.edu/faculty-resources/addressing-generative-ai/?) warns that detector scores can't conclusively identify AI-generated work; [American University](https://www.american.edu/oit/security/responsible-use-of-ai.cfm?utm_source=chatgpt.com) actually stopped using detector scores as evidence of AI generation.

In an [ongoing study from Conjointly](https://conjointly.com/blog/real-vs-ai-images-2025/), AI-generated images are correctly identified 52% of the time, while authentic photos are correctly identified 49% of the time, with human accuracy in detection at roughly the toss of a coin. So the challenges are far from over. And increasingly, [identifying where an image came from and how it was processed](https://www.adobe.com/ai/research/202609/how-an-award-winning-photojournalist-proves-photo-authenticity.html?utm_source=chatgpt.com) is becoming more and more a part of a photographer's job, and is something I've been exploring more and more myself. The [Canadian Association for Photographic Art](https://capacanada.ca/updated-ai-requirements-on-capa-competitions/?) has developed an authentication approach that's extremely close to what I'm also considering:

1. Automated screening: Run the submitted content through a detector **only as** an initial review. Record the tool, date, result, and version of the image being tested
2. Photographer-led review: Examine the image and any specific anomalies, including lighting, optical behavior, perspective, compositing, and available camera metadata; request the original file if the concern is substantial
3. Creator clarification: Ask how the content was produced and whether AI-assisted or other elements were incorporated; allow explanations and supporting evidence before asserting misconduct
4. Written assessment: Produce a brief report documenting observable facts, supporting evidence, uncertainty, and specific concerns regarding the image/content
5. Review: The report serves as a company-level document exploring all facets of alleged AI usage, including provenance, disclosure of usage, etc. This provides a more comprehensive approach to deciding whether published content violates company/group ethics, whether further information should be requested, or whether the case should be closed without action.

If you're using images professionally, even if you're not a "professional photographer", I'd suggest preserving the following: 

- Untouched camera-original JPEGs and any available capture metadata.
- Neighboring images from the same shoot when practical.
- Editing history and exported deliverables for significant projects.
- Client briefs, production notes, or other ordinary records that help establish the work's history.

However, it's important to distinguish between questioning the authenticity of an image and accusing someone of misrepresenting their work. There is nothing inherently wrong with asking how an image was created, particularly when accuracy or authenticity matters. But a concern raised by an AI detector relies on an unreliable tool that multiple institutions reject in there own quest for ensuring authentic work. So it's essential to examine available evidence, ask questions, and allow the creator to explain their work - not just to identify AI-generated imagery, but to protect the integrity of the original photography as well as the photographer and the institutions, companies and groups using the photos.

In this day and age, asking about the source of an image is prudent, especially in cases where AI may be deemed inappropriate for whatever reason. But having a clear review process for companies, groups, contests, etc. helps to preserve the integrity of published content on all sides. For photographers, preserving original images and their associated capture data provides a stronger foundation for establishing how their work was created. While this information may not answer every question about editing or AI involvement, it can provide meaningful evidence of photographic authenticity, ensuring that questions about a creator's honesty are evaluated fairly rather than left to the judgment of an automated detector.

![Discovery Call](https://flyboyz.net/hs-fs/hubfs/Discovery%20Call.jpg?width=250&height=250&name=Discovery%20Call.jpg "Discovery Call")

[Book a Discovery Call!](https://calendar.app.google/XsYwuVCE58PPuWQBA)

\*Fly Boyz LLC reserves the right to cancel all flights on the basis of Safety, Legality, and/or Weather

\*All Services are offered with no travel fees throughout most of Albuquerque, Tijeras, Edgewood and Moriarty. Services are offered throughout NM, with a nominal travel fee depending on distance. Reach out for more details.

 \*All services are subject to Fly Boyz LLC’s [**Terms and Conditions**](https://flyboyz.net/termsconditions). By purchasing or engaging Fly Boyz LLC for any services, regardless of category, size, or scope, you agree to be bound by these [Terms and Conditions](https://flyboyz.net/termsconditions). 

[facebook-f icon](https://www.facebook.com/profile.php?id=61565167644360) [linkedin-in icon](https://www.linkedin.com/in/shea-bender/) [Follow us on Facebook](https://www.youtube.com/channel/UCb-zJ7Y8XBOjopoJ6ZZbOxA)

- [Terms & Conditions](https://flyboyz.net/termsconditions)
- [Contact Us](https://flyboyz.net/contact)

- Phone: 505-318-8286
- Email: shea@flyboyz.net

[![](https://flyboyz.net/hs-fs/hubfs/Photo_Bucket__50_-removebg-preview-1.png?width=250&height=147&name=Photo_Bucket__50_-removebg-preview-1.png)](https://flyboyz.net)

 

**Copyright © 2025 Bender Aerospace LLC**

---

```json
{
  "@context" : "https://schema.org",
  "@type" : "BlogPosting",
  "author" : {
    "@type" : "Person",
    "name" : "Shea Bender",
    "url" : "https://flyboyz.net/fly-boyz-llc-blog/author/shea-bender"
  },
  "dateModified" : "2026-11-01T22:21:21.451Z",
  "datePublished" : "2026-11-01T22:21:21.000Z",
  "headline" : "Photographic Integrity",
  "image" : [ "https://flyboyz.net/hubfs/20261101_151525.jpg" ],
  "mainEntityOfPage" : {
    "@id" : "https://flyboyz.net/fly-boyz-llc-blog/photographic-integrity",
    "@type" : "WebPage"
  },
  "publisher" : {
    "@type" : "Organization",
    "logo" : {
      "@type" : "ImageObject",
      "url" : "https://flyboyz.net/hubfs/Photo_Bucket__50_-removebg-preview-1.png"
    },
    "name" : "Fly Boyz LLC"
  }
}
```