Fighting AI Fake Photos! iPhone Camera App "ZCAM" Launches, Reclaiming Reality Through Encryption Technology

To combat AI forgery, a new startup has launched an iPhone camera app called “ZCAM.” Using hardware security isolation zones and the C2PA standard, it writes a signature at the moment of capture to ensure the image has not been tampered with, helping the public independently verify authenticity.

Introducing the ZCAM App, focused on countering AI fake images

Cryptography startup Succinct Labs, supported by venture capital firm Paradigm, has recently launched an iPhone camera app called “ZCAM,” with the goal of addressing the threats posed by AI-generated photos and videos.

Succinct Labs says that at the moment an image is captured, the ZCAM App adds a cryptographic signature to the image, creating an immutable record that directly links the content to the recording device. With this technology, the public can independently verify whether media files come from genuine devices and confirm that the images have not been digitally modified or generated by AI.

Although commercial AI detection tools are already available on the market, Succinct Labs tested seven well-known ones. While these tools perform reasonably well when handling unmodified images, once the images undergo simple edits such as blurring, compression, or adding noise, the detection rate drops by 96% significantly. That is why the research team wants to use encryption technology to build products that can truly distinguish real from fake.

Image source: Photos taken with ZCAM App include detailed hash values and other data

Users can already download “ZCAM” from the AppStore, take a photo, share it, and prove authenticity at https://zcam.succinct.xyz/.

Combining two major standards to ensure images are not tampered with

The way ZCAM works is that when a user takes a photo or records a video through the ZCAM App on their iPhone, the system first calculates a cryptographic hash value of the original pixels. Next, the iPhone uses its built-in hardware security enclave (Secure Enclave) to generate a dedicated private key to sign that hash value.

Secure Enclave is a tamper-resistant co-processor. Key generation and storage are carried out in an isolated environment, meaning the key never leaves the iPhone hardware. After that, Apple’s App Attest service generates a proof to ensure that the signature indeed comes from the ZCAM application itself.

These capture metadata, signatures, and proofs are embedded into the media file as a C2PA manifest.

Image source: Explanation of ZCAM App’s technical principles

C2PA is an open standard jointly developed by organizations including Adobe, Microsoft, Google, OpenAI, and the BBC. When a platform receives a photo or video captured by ZCAM, it extracts the C2PA manifest from the file, recalculates the hash value, and checks the signature.

As long as any pixel changes, the hash value will not match. If it matches perfectly, viewers can confirm that the photo is a genuine image.

AI scam losses projected to reach $40 billion in 2027

As generative AI technology becomes increasingly realistic, the associated scam risks are also rising.

Succinct Labs cites forecast data from Deloitte, saying that by 2027, generative AI will likely lead to U.S. scam losses reaching $40 billion, a significant increase from $12.3 billion in 2023.

Since commercial AI detection tools are likely to fail, directly using smartphone hardware to generate unique cryptographic signatures would be a more effective approach.

Although many people in the industry believe that proving something is authentic is a better solution than simply detecting fakes, the biggest challenge for the ZCAM App at present is how to incentivize widespread real-world usage and further expand user adoption.

In a statement, Succinct Labs emphasized the commercial application potential of ZCAM and expects the technology to bring tangible benefits to enterprises and news professionals.

Paradigm leads $55 million investment to build a zero-knowledge proof network

The ZCAM App development team, Succinct Labs, has a strong foundation in blockchain and cryptography, and they completed a $55 million funding round back in 2024, led by venture capital firm Paradigm. Participants also included the founders of well-known crypto projects such as Polygon and EigenLayer.

Succinct Labs says that its product, SP1 zero-knowledge virtual machine (zkVM), currently protects more than $4 billion in crypto assets.

Then, in August 2025, Succinct Labs also launched the mainnet of the Succinct Prover Network and simultaneously activated its native token $PROVE. The network provides a decentralized marketplace on Ethereum, enabling various dApps to submit zero-knowledge proof request, which is then verified through competition by independent provers.

Further reading:
ZK technology goes mainstream! Google Wallet adopts zero-knowledge proofs to create your digital ID. Understand zero-knowledge proofs! Explain the working principle of zkSync circuits using the “zoo” concept