Starling Lab: Building Trust in Human Data
Today, USC Holocaust Foundation and Stanford University jointly unveiled Starling Lab, a new research center dedicated to solving technical and ethical challenges using the latest advances in cryptography and decentralized network protocols, Build to build trust in the most sensitive digital record in human history.
Protocol Labs and the Filecoin Foundation initially committed $2 million in funding for the decentralized web, and Starling Labs is the world's first center dedicated to using decentralized tools to advance human rights.
Nearly three years after Protocol Labs began collaborating with researchers at Stanford and USC, today’s announcement: Together, we have launched the most comprehensive set of tools and principles to advance human rights through distributed network technologies.
However, the world has changed since we started. Current events expose the importance of trust to social justice and scientific progress. From the fight to end systemic racism to promoting fair and democratic elections and limiting the spread of COVID-19, digital documents fuel these movements.
It is clear that in an era of declining trust in digital media, the vulnerability of digital documents remains. Bits alone cannot solve this crisis. The human rights community recognizes that it needs new tools and values to address these challenges. While many are just a step away, what is needed is a new generation of influential labs developing the tools and education for their success.
If you're interested in working with Starling Lab to capture, store, validate, and preserve valuable datasets, visit starlinglab.org/join to learn more about training, education, and funding opportunities.
online disinformation
Digital information comes in a flood: information can be captured by anyone with a smartphone or camera, and limited tools exist to prevent individuals from altering captured data or misattributing it before they choose to share it.
Ease of manipulating data: A host of artificial intelligence tools, "deepfake" techniques and other applications make it easy for parties to alter written, photographic and video data, as well as obscure records of altercations.
Verifying data is difficult: Current Internet solutions do not provide an easy and comprehensive way for individuals to verify the integrity of data globally in a sustainable and trustworthy manner.
Starling Lab
Starling Lab collaborates with USC Holocaust Foundation, Stanford University, and deploys technologies and methods to make the decentralized web a viable platform for social impact. Starling Lab is designed to meet the following needs:
1. Provide individuals with a way to access crowdsourced data in order to establish a chain of custody from the start;
2. Store data in a decentralized manner and cannot be manipulated or changed;
3. Provide the ability to verify data without trust.
At the heart of Starling Lab's solutions is the Starling Framework, a set of open-source prototypes, best practices, and case studies that help reduce information uncertainty in digital media. The Starling framework is based on the fundamental need to fetch, store, and verify information to protect its integrity.
fetch information
In Capture, a combination of hardware (HTC) and software (IPFS and Filecoin) technologies has been prototyped to create a chain of custody from camera to digital platform. The images are paired with metadata from an array of sensors on the device to prove that the footage was taken at a specific time, date and location. All of this material is then cryptographically decentralized, creating a Content Identifier (CID) that acts as a unique fingerprint for that material.
shop
In the Store, data is replicated to decentralized storage nodes, such as IPFS and Filecoin, which natively use content identifiers (CIDs) for addressability. If a single pixel is changed, the encryption algorithm will generate a completely different hash value for the footage.
verify
When it comes to validation, to handle all hashes generated during fetching and storage, the Starling Framework has a hash/authentication management system that allows organizations to hire multiple experts to validate footage.
Starling Labs has deployed the Starling data integrity framework in three major case studies:
Starling Labs has deployed the Starling data integrity framework in three major case studies:
Protecting the Election Coverage Showcase
A team from Starling Lab and Reuters deployed the Starling framework during the 2020 California primary to explore how cryptography could enable and protect the work of its photojournalists in the age of misinformation.
Reuters reporters scoured photo prototypes to determine their provenance and make them more public trustworthy. In addition to Canon cameras, Reuters also uses the Starling mobile app, which allows journalists to encrypt metadata, including barometer, gyroscope, GPS data, and send all the information to Reuters. The result was a barrage of information confirming that the photo was taken at that location, and at exactly the time Reuters said it was.
78 Days: Create a Trusted Photography Profile
Over a period of 78 days, a team from Starling Lab and Reuters used a series of new image authentication technologies and decentralized network protocols to document the transition of the US president from Donald Trump to Joe Biden.
Using the Starling framework and the new Content Authenticity Initiative's attribution standard, Starling Lab created a prototype archive that allows each photo to become a container for image pixels and additional metadata and links.
Genocide Evidence Showcase
A team from Starling Lab and the USC Holocaust Foundation deployed the Starling framework to cryptographically capture testimonies of survivors of the Holocaust, the Armenian genocide, and the genocide in the Rohingya crisis.
To meet these needs, the team first uploaded all current USC Holocaust Foundation data to IPFS and Filecoin so that it can be secured through on-chain decentralization. As more people engage with and contribute to stored content, file archiving becomes more secure and history becomes more resilient.
Next, the team used the Starling framework to scrape new interviews with survivors. Using phones and DSLRs, the team was able to scrape, store, and authenticate content from the start, securing interview data and metadata and storing it securely on IPFS and Filecoin. All of these videos have been added to the USC Holocaust Foundation archives.
