How the AI Executive Order will Accelerate the Convergence of AI and Blockchain

President Biden recently issued an AI Executive Order to manage the risks of and seize the opportunity presented by artificial intelligence.

This is the most sweeping set of standards to increase safety and security of AI, protect citizens’ privacy, advance equity and civil rights, and promote innovation. The ultimate goal of the EO is to ensure safe, secure, and trustworthy use of artificial intelligence.

Deep Ventures is an active investor in both AI and blockchain and we believe that there are many synergies that arise from the merging of these technologies. We see two very important initiatives laid out in Biden’s EO that may accelerate this convergence.

Blockchain can help protect users from fraud and deception caused by AI

One of the key actions of the EO is to “protect Americans from AI-enabled fraud and deception by establishing standards and best practices for detecting AI-generated content and authenticating official content.”

Biden has assigned responsibility to The Department of Commerce to develop guidance for content authentication and watermarking to clearly label AI-generated content. Federal agencies will follow these guidelines and use tools to ensure that Americans know that the communications coming from the US government are legitimate, with hopes that the private sector and other governments will follow suit.

AI can now reproduce people’s voices and facial movements with astounding accuracy, and it’s can be very difficult to decipher what is real or fake. This may lead to counterfeit content that seems to come from government officials. If this content is provocative and offensive to other nations, it can literally start world wars.

Blockchain can help identify images, videos, and other media as real or fake.

For example, communications that come from The White House, such as the AI Executive Order itself, can be cryptographically signed by the Press Secretary and other team members and put on the blockchain for anyone to verify. All nodes on the blockchain (as opposed to a single centralized party) need to come to consensus, providing a much higher level of confidence that the content is authentic.

On a call with reporters, a White House representative actually said, “I can imagine, honestly, a version of a call like this in some number of years from now and there'll be a cryptographic signature attached to it that you know you’re actually speaking to [the White House press team] and not an AI version.”

Deep Ventures’ portfolio company Numbers Protocol is working on exactly this problem, and has worked with influential media companies such as Reuters, Rolling Stone, and others to enhance content provenance.

Blockchain can help preserve the privacy of data used in AI models

Data is at the core of AI, as a plethora of data is necessary to train Large Language Models (LLMs) and other AI schemas effectively. We’ve seen the efficiency benefits that these AI models provide, but the way the model data is extracted, analyzed, and integrated must be done in a way that maintains citizens’ data privacy.

Blockchain can play a huge role in ensuring that the data used in AI models stays private. Here is a small sample how:

1. Secure identity verification and authentication - Blockchain can be used to establish and verify decentralized digital identities, reducing reliance on centralized identity providers and single points of failure. Smart contracts can then be used to authenticate who has permission to access and modify specific data for use in AI models.

2. Data encryption and decentralization - Blockchain technology offers a secure, immutable, and decentralized ledger for storing data. Compared to a centralized server, this makes it more difficult for hackers to access the data, thus improving security.

3. Data anonymization - Privacy methods such as homomorphic encryption, federated learning, and differential privacy can be applied to analyze and extract insights from data without revealing sensitive information about individuals. By integrating these privacy-preserving methods with blockchain, data can be securely shared and analyzed without compromising privacy.

4. Data ownership and consent

   1. In conjunction with DIDs, blockchain allows users to have more control over their personal data and give them the ability to provide and revoke data access. Additionally, the cryptocurrencies associated with blockchain networks can provide financial incentives for users to provide access to their data to train AI models.

5. Auditing and accountability - Blockchain's transparency and immutability make it easier to track and audit data access and usage. This can also help with compliance with data protection regulations such as GDPR.

We are in the midst of a privacy renaissance, and blockchain

Conclusion

The AI Executive Order lays out initiatives to ensure the safe, secure, and trustworthy use of artificial intelligence while maintaining innovation.

Blockchain will play a very important role in achieving these initiatives, namely in 1) content provenance to protect users from fraud, and 2) privacy to protect user data.

Many more use cases at the convergence of AI and blockchain will arise, and we look forward to funding and supporting companies building at this intersection in the near future.

Thank you for reading this Deep Thought. If you enjoyed this post, please share it!