The Interledger Community 🌱

Interledger Foundation Programs AI Policies and Guidelines

These guidelines define how applicants and grantees may use various artificial intelligence (AI) tools and services when preparing grant applications, grant progress reports, and other Interledger Foundation programmatic outputs. Please note that policies beyond what is provided here and specific to an individual grant program will be detailed in that program’s Grant Application Guidelines.

Our views on AI and its use

We recognize that artificial intelligence tools can be useful in professional work and are likely to be used in preparing documents, applications, and other materials for the Interledger Foundation.

Given this, we permit use of AI tools. These guidelines are designed to assist potential and active grantees in using these tools effectively, while also understanding how they may influence the Interledger Foundation program team's assessment, decision-making, and overall evaluation of submitted work.

Caution about using AI

We noticed an increase in AI-generated content in grant applications, reports, blog posts, and other assets, where writing was almost or completely done by AI. We are becoming more able to identify AI-generated content and have made it a priority to better understand its use.

We value original thought, creative thinking, and detailed specificity. AI tools can be helpful with presenting ideas in a compelling manner, but we will view them negatively when we consider these values secondary to AI-generated contributions. At the Interledger Foundation, we believe the prevalence of AI raises the bar for compelling ideas and proposals to stand out among generic auto-generated writing.

Generative AI often suffers from the GIGO Effect (Garbage In, Garbage Out). The quality of output is fundamentally determined by the quality of input. Outdated and inaccurate data used to train these AI models (Garbage In) can result in illogical or misleading statements (Garbage Out).

Important Notes

  • If you use AI, you must understand the limits of its usefulness. For spelling, grammar, and structuring a document, it can be a useful writing tool. However, current models have demonstrated limitations and poor results when aggregating data, fact-checking, providing an opinion, and citing sources.

  • Generative AI may not describe Interledger technologies accurately. Similarly named ideas and the evolving nature of technology can result in conflated or outdated misrepresentations of concepts. Inaccurate or outdated references to Interledger technologies reduce the competitiveness of an application. Review the documentation in our Developers Portal to learn more about how Interledger technologies work.

  • Be cautious when using AI to inform us about your alignment with our grant programs or foundation mission.

  • Do not use AI to generate your budget. Your project budget is a crucial part of our assessment. Your costs should be accurate and align with the activities of your project.

  • Successful applications provide specific details about how your skills and past experiences will allow you to deliver on your project.

  • Use caution when copying and pasting, as unintentional information, watermarks, or other identifiers can be included and serve as a clear indication that the work is not the author's.

Legal and ethical concerns about using AI

  • There are various ethical and legal concerns associated with using AI that you should consider when employing it for Interledger Foundation-related work, and which we are also concerned about as the industry evolves. Some of these to consider:

  • Using AI tools may involve irrevocably agreeing to the use of your work for training future models that could be used by others. Make sure you are not inadvertently giving access to confidential information, your intellectual property, and other valuable project/organization/company assets.

  • There are numerous debates, laws, and controversies surrounding the legal use of copyrighted material for training AI models. Are you adding to this problem by using AI tools? Are you exploiting the uncredited work of others? Could the use of AI put you in legal jeopardy of plagiarizing others? These are issues you should think about as you decide to use it.

  • AI relies on extensive “data work” (e.g., data collection, cleaning, annotation, and evaluation). This work is often outsourced to precariously employed contractors, frequently in the Global South, who face low pay, unstable hours, and mental health risks from moderating or labeling harmful material. Multiple independent investigations and labor studies have documented these harms, and they continue to be an active human rights concern in AI supply chains. The Interledger Foundation will not knowingly fund projects that benefit from exploitative data labor practices or non-consensual data harvesting.

  • The detrimental use of AI in our global environment is clear and stark. For a comprehensive report on the environmental issues with AI, read The Green Web Foundation’s report, Thinking About Using AI?

  • The carbon emission platform Arbor describes the effects this way:

    • AI's Environmental Impact: The growth of AI significantly increases energy use, contributing to global emissions comparable to small nations.
    • Data Center Demand: AI's rise has led to a 72% increase in data center power consumption from 2019-2023 alone, significantly contributing to emissions.
    • AI Query Emissions: Each interaction with AI systems like ChatGPT results in 4.32 grams of CO₂e, highlighting the need for emission reduction strategies.
    • Industry Initiatives: Companies like OpenAI and Google are adopting renewable energy and carbon-neutral operations to curb AI's environmental impact.‍
    • Sustainability Strategies: Reducing AI’s footprint involves adopting energy-efficient technologies, using renewables, optimizing data centers, and managing e-waste responsibly.

Disclosure Requirement

Applicants who use AI tools in preparing their application must disclose on the application form which categories of use (for example, idea generation, grammar editing, literature review, translation, graphics design, writing responses) and which tools they have used for generative AI in the preparation of their applications. We may request your prompts. We suggest using the Badge AI notation for disclosing the genesis of your work.