Succinct Labs is stepping into the growing battle against AI-generated misinformation with a new product designed to prove ‘what’s real’ at the point of creation.
ZCAM to detect AI frauds using cryptography
Paradigm-backed Succinct Labs introduced a new product called ZCAM on Thursday. The product is an iPhone camera app, which would rely on cryptography to measure the authenticity of videos and photos as and when the image is captured.
The company said that ZCAM “signs photos and videos at the moment of capture, producing a tamper-proof record that links content to the device that captured it.”
ZCAM focuses on establishing a trusted record from the start. They aim to combat the major concerns of AI-generated images and videos used in online frauds, identity abuse, and concerns around false media.
Succinct stated, it “signs photos and videos at the moment of capture, producing a tamper-proof record that links content to the device that captured it.” They added that individuals can independently verify the images of videos to find out if it came from a real device or if it has any AI alterations.
Using Cryptography to Prove What’s Real
Succint is an ‘applied cryptography company’ with the fastest zero-knowledge virtual machine named SP1, securing above $4 billion in digital assets, according to the report.
According to the company, the technology helps in proving transactions, verifying computation, and eliminating the need for intermediaries in securing assets.
Succinct’s research pointed at how commercial AI detectors are more prone to fail. The company’s approach is to focus on the hardware. This way even smartphones can “generate unique cryptographic signatures.”
Whenever a user takes an image or video using an iPhone, ZCAM generates a cryptographic hash from the captured pixels, Succinct said.
Is ZCAM the first of its kind?
The technology is not new. Projects like World, backed by OpenAI CEO Sam Altman, have also been using blockchain technology to mitigate the risk of AI.
World, the real human network, as reported, has been built to ensure the separation of real people from AI-driven online accounts. It was developed in response to the concerns of increasing bot accounts on social media platforms.
The platform assigns user IDs to people after they prove ‘they are human,’ a strategy designed to allow for distinguishing AI-powered actors from real people online.
However, both World and Succinct are working towards building authenticity. ZCAM builds on that broader foundation, applying cryptographic verification to one of the most visible challenges of the AI era to reinstate trust in digital media.
AI fraud risk rise, as per report
Financial service platform Deloitte had published a blog emphasizing the role of Generative AI in magnifying fraud in the banking industry, with a figure of $40 billion in loss reported in the U.S. by 2027.
“Governments, companies, and researchers have invested heavily in tools trained to detect AI-generated content,” Succinct noted. However, the company believes that these efforts are well-intentioned, but they do not work.
Last August, Succinct launched the mainnet for its Succinct Prover Network alongside its PROVE token. The network operates as a decentralized marketplace on Ethereum, where applications can submit zero-knowledge proof requests and independent provers compete to validate them.
The Prove network is providing a decentralized marketplace on Ethereum. This way, the applications can request zero-knowledge proofs and independent provers competing to validate their requests.
