Blockchain service provider Casper Labs has announced plans to integrate artificial intelligence (AI) systems with Web3 technologies for improved transparency and excellent consumer safety.
Casper Labs said it will collaborate with technology firm IBM (NASDAQ: IBM) to combine emerging technologies for new use cases, including developing a solution to regulate AI training governance across several organizations.
The partnership will seek to create a policy enforcement layer for firms, ensuring that organizations deploying generative AI models conform with the original model’s design. Casper Labs will rely on IBM’s Watson. Governance and publicly distributed ledger to achieve this primary objective.
In deploying generative AI models, it is not uncommon for firms to modify large language models (LLMs) by introducing new data sets in ways that may be at variance with the original model training. Creators are often unaware of the wholesale changes employed by firms deploying their models, leading to intellectual property (IP) issues.
Casper Labs says its blockchain-based solution will allow organizations to share information on its ledger concerning the changes made to AI systems. The firm added that exchanging information reduced the risk of IP crossover while allowing creators to “remediate issues” that may crop up during deployment.
“With IBM’s help, we’re committed to delivering a better way to not only understand why AI systems behave the way they do but also a clearer path to remediate behavior if hallucinations or performance issues occur,” Casper Labs CEO Mrinal Manohar.
Casper Labs’ solution will rely on a compliance dashboard, version control to roll back to previous versions, a quality control toolkit, and an audit and reporting system.
The blockchain service provider is eyeing use cases across financial services, retail, healthcare, and manufacturing, with many firms pivoting to generative AI to improve their productivity.
“AI’s long-term potential will be dictated by how effectively and efficiently organizations can understand, govern, and react to increasingly massive AI training data sets,” said Manohar.
The marriage of AI and blockchain
As copyright and data handling issues continue to threaten AI’s long-term future, experts are pushing for a wholesale integration with blockchain as a solution. Industry stakeholders point to blockchain’s tamper-proof and transparency properties as features that could improve AI platform data handling and safety.
Others are exploring new use cases, including using blockchain to track content usage by AI firms to prevent copyright infringement.
“Given that blockchain is tamper-resistant, completely time-synched, and can be completely automated, AI governance can be done right, be highly automatable, and highly certifiable right out the gate,” Manohar said.
While AI is regarded as a critical technology that will supercharge innovation, some are still adamant about integrating it into their services despite its growing adoption in enterprises, likely rooted in the tech’s infringement and fair use issues.
This is where the importance of integrating AI with blockchain comes in.
Giovanni Franzese, former head of blockchain at Ericsson, said blockchain provides data security that AI lacks in its current state.
“AI will be another key technology to combine with blockchain, and we will see the wonderful things that the technologies will manage to achieve,” noted Franzese in an earlier interview with CoinGeek. “AI needs to achieve the traceability and immutability of data, which only blockchain will achieve.”
Besides providing data security, blockchain also validates data authenticity, which is essential for AI developers, who rely heavily on data to train their large language models (LLMs), according to nChain Director of Commercial and Strategy Simit Naik.