For most of its history, a photograph was strong evidence that something happened. Today, AI can generate realistic images of events that never occurred and people who never existed, and can alter real images convincingly. This affects journalism, courts, insurance, science, politics and personal lives. This lesson explains what synthetic media is, how it changes the value of photographic evidence, how professionals verify images, and what responsibilities photographers have in this new environment.
What is synthetic media?
Synthetic media is content wholly or partly created or altered by AI to look realistic, including generated images, face swaps, voice clones and video deepfakes. It ranges from openly creative work to deliberate deception. Image integrity principles were covered in truth, manipulation and image integrity, and provenance tools in image authenticity and provenance.
| Type | Example | Typical risk |
|---|---|---|
| Fully generated images | A realistic street scene that never existed | False news or fake evidence |
| Generative edits to real photos | Adding smoke, crowds or objects | Misleading context |
| Face swaps and deepfakes | A person's face placed into another image or video | Harassment, fraud, reputational harm |
| Synthetic people | Realistic portraits of non-existent people | Fake profiles and scams |
How evidence is affected
- False images can spread quickly before they are checked.
- The "liar's dividend": a term used by legal scholars for the way real evidence can be dismissed as fake simply because fakes exist.
- Verification takes longer for newsrooms, courts and insurers.
- Trust shifts toward sources and provenance rather than the image alone.
Tip:As fakes become easier to make, a trusted photographer's reputation and documented process become more valuable, not less.
How professionals verify images
| Method | What it checks |
|---|---|
| Source and contact | Who took it, can they be contacted, do they have other images from the scene? |
| Reverse image search | Has the image appeared before, elsewhere or earlier? |
| Original files and metadata | Camera files, capture times, sequences, as in technical metadata and reporting |
| Content Credentials and watermarks | Signed history or AI markers, where present |
| Visual and contextual analysis | Shadows, weather, signs, landmarks and physical consistency |
| Corroboration | Other independent images, witnesses or reports |
No single method is decisive. Professional verification combines several, as also outlined in critical thinking and visual literacy.
Rules and labelling
Governments, platforms, news organisations and competitions are introducing rules on synthetic media, such as requirements to label AI-generated or AI-altered content, restrictions on deceptive political content and bans on generative AI in news and documentary categories. Rules differ and change quickly, so check the requirements of each place you publish and each client or competition you work with.
Photographers' responsibilities
- Never create deceptive synthetic images of real people or real events.
- Label generative content clearly where viewers might assume reality.
- Keep originals and process records for documentary and commercial work.
- Use provenance tools where available.
- Protect subjects: be careful how and where images of people are shared, to reduce misuse in deepfakes.
- Report harmful deepfakes using platform tools when you encounter them.
Warning:Creating or sharing realistic fake images of real people, especially intimate or defamatory content, can cause serious harm and may be illegal in many places.
Common mistakes
- Sharing dramatic images before checking: spreading falsehoods.
- Assuming realistic means real: synthetic images can look perfect.
- Assuming every surprising image is fake: dismissing genuine evidence.
- Unlabelled creative AI work: confusion and lost trust.
How professionals respond
Newsrooms, documentary photographers and responsible brands invest in verification, provenance and transparent labelling. Photographers protect their credibility by keeping originals, documenting their process, refusing deceptive work and being open about AI use. In a world of synthetic images, honesty becomes a competitive advantage.
Practical examples
A dramatic image after a storm
- Situation
- A striking photo of flooded streets spreads online after a storm.
- What to do
- A newsroom contacts the source, requests original files, checks earlier appearances and compares with other reports before publishing.
- Why it works
- Combined verification protects against fakes and recycled images.
- Result
- The editors publish only images they can confirm.
A creative AI series
- Situation
- You create a surreal AI-assisted series about future cities.
- What to do
- You label the series as AI-assisted digital art and include Content Credentials.
- Why it works
- Clear labelling prevents viewers mistaking it for documentary photography.
- Result
- Audiences appreciate the creativity without being misled.
Key points
- Synthetic media includes generated images, generative edits, face swaps, deepfakes and synthetic people.
- It allows convincing fakes and lets real evidence be dismissed as fake (the liar's dividend).
- Verification combines source contact, reverse search, originals, provenance, analysis and corroboration.
- Labelling rules and AI restrictions are growing and vary; check each context.
- Photographers must avoid deception, label generative work, keep originals and protect subjects.
Frequently asked questions
What is synthetic media?
What is the liar's dividend?
How can I tell if a photo is AI-generated?
Do I need to label AI-generated images?
Conclusion
Synthetic media challenges the trust placed in photographs. Understand its forms and risks, verify carefully, follow labelling rules, refuse deception and protect people, so your images remain trustworthy. Next, you will learn to match technology choices to real photographic needs.
Finished reading? Track your progress through AI and new technology.