Research Summary: “Blockchain Ethics Research: A Conceptual Model”

TL;DR

  • Propose a broad framework for studying ethics in the cryptocurrency and smart contract space.
  • Identify 4 major blockchain “realms” that ethics apply to:
    (1) Technology stack - ethics related to underlying distributed ledger technology;
    (2) Cryptocurrencies - ethics of cryptocurrencies as an alternative to fiat;
    (3) Smart contracts - ethics related to the design and implementation of smart contracts and data;
    (4) Decentralization - ethics related to the design of decentralized autonomous institutions (DAOs).
  • Areas relevant to smart contracts and DONs revolve around data ethics and contractual ethics which are discussed below.

Citation

Tang, Y., Xiong, J., Becerril-Arreola, R. and Iyer, L., 2019, June. Blockchain ethics research: a conceptual model. In Proceedings of the 2019 on Computers and People Research Conference (pp. 43-49).

Link

https://dl.acm.org/doi/pdf/10.1145/3322385.3322397

Core Research Questions

Which blockchain features are most relevant to discussions about ethics? What should discussions about blockchain ethics focus on?

Focused Summary

The authors identify established ethics research relevant to smart contracts and DONs: data ethics and contractual ethics.

  • Data ethics are relevant to cryptographic protocols such as DECO and Town Crier detailed in the Chainlink 2.0 whitepaper;
  • Contractual ethics apply to features of hybrid smart contracts. Each are discussed below and will be fleshed out in greater detail in subsequent posts.

Background

For background terms and definitions, check out this excellent summary of the Chainlink 2.0 whitepaper.

Cryptographic Protocols Through a Data Ethics Lens

Research on data ethics typically revolves around a framework known as PAPA (Mason, 1986) which stands for Privacy, Accuracy, Property and Accessibility. This framework provides useful heuristics for identifying potential ethical strengths and weaknesses of cryptographic protocols such as DECO, which is discussed in more technical detail in the Chainlink 2.0 whitepaper.

Below is a table with each of the areas of data ethics and a series of critical questions within each area. Using this framework, potentially as an ethics certification “checklist”, we can potentially validate cryptographic protocols as “ethically sound” if they provide sufficient answers to each of critical questions raised (this will be the topic for the next discussion post in the series).

Area Critical Questions
Privacy What data must people reveal about themselves to others?
Accuracy Who is responsible for the reliability, authenticity, and accuracy of data? Who is accountable to errors in the data?
Property Who owns the data? Who owns the channels of distribution, and how/should they be regulated? What is the fair price of data that is exchanged?
Accessibility What data does a person or organization have a right to obtain, with what protection, and under what conditions?

PAPA Framework for Data Ethics partially reposted from the Binus University School of Information Systems which can be used to assess the ethical integrity of cryptographic protocols.

Contractual Ethics and Hybrid Smart Contracts

Hybrid smart contracts touch upon two areas of ethics: contractual ethics and data ethics. While data ethics are more relevant to cryptographic protocols like DECO and Town Crier, contractual ethics are more relevant to smart contracts as a broader category of transformational technologies. Because smart contracts are based on automated rules, ethical considerations might appear to be more straightforward.

This is not necessarily the case, however, if we consider the legal implications and social norms surrounding “traditional” vs. smart contracts. While traditional contracts are validated and adjudicated through centralized courts and governments, smart contracts are validated through decentralized peer-to-peer networks and are not only binding in a legal sense but are binding via abstract, automatic processes.

As a result, ethics and fairness considerations relevant to smart contracts will likely come into play in the following ways:

  • Consent: were the rules of the smart contract designed in a way to ensure informed consent of both parties?
  • Compliance: would the smart contract be legally enforceable in the relevant jurisdictions in which the smart contract was executed?
  • Validation: are there aspects of the validation process, or of the validators themselves which could potentially void a smart contract?

These are only a few of the ways in which considerations about ethics and fairness could come into play for smart contracts. A deeper understanding of the fairness and ethics issues in this domain would require exploration of these issues via case studies with real-world examples and applications.

Thinking through some of these issues before they become politicized and/or controversial can help contribute to better smart contract design and to a fairer and more inclusive economic future.

References

Mason, R.O., 1986. Four ethical issues of the information age. MIS quarterly , pp. 5-12. Link: https://www.jstor.org/stable/pdf/248873.pdf

Up next…

Discussion: A data ethics framework for assessing cryptographic algorithms and zero knowledge proofs: PAPA, DECO and Town Crier.

8 Likes

I have researched a fair amount on AI ethics but not nearly as much on blockchain ethics, so this is a very cool topic to study. There is something fairly obvious held in common between AI ethics and blockchain ethics - it’s all about democratization of the benefits that they provide.


Our lab’s work “Multi-Layer Aggregate Verification for IoT Blockchain” is all about leveraging a multi-layer IoT blockchain structure in order for farmers to be provided equal-opportunity nodes along with financial institutions and relevant inspection committees. This is the first time I’ve read about the PAPA framework, which shows I’m enough of a noob. Anyways, I took the chance to analyze our paper under this framework, so this section of the comment will become relevant once I release this paper’s summary (I’ll subsequently place a reference to this comment).

Relevant personal information such as name and crop/land location is provided by the farmer to the bank and government.

It’s the responsibility of all nodes in the network to verify that the information provided by the farmer is accurate.

Agricultural IoT data is uploaded to the blockchain and is publicly owned (the blockchains used are Ethereum and Ethereum-based Quorum), and the data uploaded on the database is owned by the financial institution. Smart contracts regulate a data reviewing process.

As mentioned above, all three node types have access to all blockchain data. I do have a question on this: wouldn’t the privacy of the data necessarily determine its accessibility? Where is the distinguishment between the two?

Regarding what you wrote about contractual ethics:
Consent: the smart contract is not instantiated unless the financial institution and farmer may reach an agreement on the crop yield required
Compliance: if the farmer fails to meet the required crop yield, the bank will deny access to a loan. If the farmer successfully meets the crop yield, the financial institution must provide the loan based on the terms of the contract.
Validation: actions that may invalidate the smart contract could be made through workarounds in our proposed IoT system, for example if data is somehow falsified.

7 Likes

Thanks for this awesome reply. To answer your question about privacy, I agree that privacy and accessibility will be directly correlated in many contexts but how they interact with one another will depend upon who the ethics framework is being applied to and requirements to divulge information.

For instance, in a followup discussion piece on zero knowledge proofs that I’ll post soon, the “prover” has full accessibility to their data as well as full privacy while the “verifier” has no accessibility. However, in another context, a prover can have full privacy rights as well as accessibility but may choose to divulge information for strategic reasons or to receive some benefit. In the latter case privacy and accessibility would be the same for the prover, but different for the verifier.

Unfortunately, the PAPA framework is a bit mushy and ambiguous because it was created to identify issues and generate discussion rather than make normative prescriptions.

This paper does a good job of spelling out some of the limitations of the PAPA framework:

Woodward, B., Martin, N.L. and Imboden, T., 2011. Expansion and Validation of the PAPA Framework. Information Systems Education Journal , 9 (3), p.28.
Vancouver

Over the next three posts or so I will try to narrow down some of the parameters of the framework for zero knowledge proofs, DECOs and decentralized identity. At least that’s the plan! Thanks again for the comment.

4 Likes