Skip to main content
x
APIToS Site survey

Building sustainable infrastructures? Participate in the APIToS (API Terms of Service) survey

"I have read and accept the terms of service" is the biggest lie on the web. This is the motto of the "Terms of Service; Didn't Read" (ToS;DR) project started in 2012 to help make Terms of Service (ToS) more readable [1].

Ten years later, and despite increased consumer protection, this topic is still not well understood and it is still quite common for service users to click on "I have read and accept the terms and conditions" without having read or really understood the content of the underlying contract. The interconnection of new digital services makes these concerns even more complex and crucial, because they condition access to one or more APIs (Application Programming Interface), which can be considered the cornerstones of our digital infrastructure and economy.

Today, the APIToS (Digital Infrastructure Grant) research project aims to explore ways to make API service conditions more sustainable, accessible, understandable, and usable by humans and machines. To do so, we are collecting your experiences and opinions through a short questionnaire (15 minutes) entitled "APIToS: towards a trustworthy and sustainable ecosystem". As users, suppliers, or actors involved in the regulation of APIs, your feedback will be very helpful to us.

Please visit our website apitos.org and take part in our survey before March 30th, 2022.

To learn more about the APIToS project, the preliminary results and the objectives of the questionnaire, read the rest of the article below.

---

API Terms of Use: Updates with Massive Implications

APIs are the foundation of our digital economy based on data and the creation of new services for its collection, analysis and commercial exploitation. The access and durability of these interfaces are now recognized as necessary in all fields (finance, research, transport, etc.) and are even added to the European requirements in terms of Open Data (open data directive with regard to high value data). API Terms of Services (ToS) are a key element in building sustainable, equitable and durable digital infrastructures. Indeed, each time you use a Software-as-a-Service via its API, you integrate the technical contract represented by the interface itself as well as the commercial and legal contract (rights to use the API, commitments of its supplier, ownership of the associated data, etc.).

However, several points of concern need to be highlighted:

  • Thousands of services are developed on the basis of Google, Twitter, Facebook, or Linkedin's APIs. Sometimes, when updates are made to the general conditions of use, it causes confusion in digital ecosystems. These changes disrupt the business models of established services (changing SLAs, data reuse conditions, etc.) and often kill the trust needed between different stakeholders (users, providers, intermediaries, etc.) to build a healthy and resilient digital infrastructure [2].
  • In other domains, such as academic research, scientific work is increasingly dependent on APIs to access and analyze data or publications through text and data mining. Although there has been a huge movement towards open access for publications and data in research thanks to the open science movement, APIs from private publishers could become a new closed or controlled door slowing down the free flow of knowledge [3].
  • APIs are also a major issue for regulators to support a fair and balanced ecosystem for different stakeholders and to empower those who are more vulnerable. Indeed, contracts related to APIs are as much a risk for consumers (lock-in, submission to abnormal conditions, etc.) as for free competition.


The APIToS project: towards a sustainable and trustworthy API ecosystem

APITOS Website

                                                                                                             Visit the apitos.org website

The first exploratory phase of the APIToS study, financed by the Ford Foundation Digital infrastructure grant, allowed us to understand in detail the stumbling blocks concerning ToS and their utility for users, suppliers and regulators : Are they read? Which clauses seem critical for users? What are the consequences of frequent changes to ToS? What regulations are possible today?

Often little read or just skimmed over, the general Terms of use represent today a thin zone of possible exchange between users and API providers and often reflect asymmetric interactions:

  • On the user side, terms and conditions often remain unfamiliar with a lack of support to help with their legal understanding. Their frequent updates make them difficult to read and can be critical depending on the modified clauses (use of data, guaranteed level of services, etc.). Moreover, even accessing APIs is sometimes a difficult task, as it can be the case in academic research where requests to use APIs (submission of applications to Facebook, Twitter, etc.) are refused without explanation.
  • For API providers, apart from the large platforms that are well equipped in terms of legal services, ToS often remain complex to draft for small and medium-sized companies developing services (on or from APIs).

Within the APIToS project, better understanding these frictions is a way to act on these pivotal spaces to build more balanced relationships between different stakeholders in complex ecosystems of services, platforms and infrastructures. What remains to be done is to construct the building blocks necessary for this ambition.

To do so, we have been interested more specifically in the possible ways of standardizing terms and conditions of use to make them more easily understandable and readable. At the crossroads between law and design (legal design), we explored various innovative initiatives concerning APIs (API Rating Agency which measures the performance of different APIs and their underlying services, API Terms of Service Generator in the Swedish legal framework) or general terms of use (like ToS;DR or the recent French governmental initiative Open Terms Archive to follow the evolution of terms of service). The Creative Commons model was also a major source of inspiration for the project to think about the implementation of standard API models that are easily understandable by anyone but also machine-readable.

The interviews conducted in the first phase of our study, in addition to highlighting the points of concern mentioned above (asymmetrical relationships, lack of communication, difficulty in reading ToS, etc.), gave us the opportunity to refine some possible solutions to facilitate their readability and their role as a contract of "trust" rather than distrust. For the second phase of the project, we invite you to give your opinion through a short questionnaire.


Survey: share your practices and your opinion

The questionnaire will take 10 to 15 minutes and is intended for API users, suppliers and other actors involved in their regulation. The objective is to better understand your use of ToS and to collect your opinion on the avenues explored to make the APIs Terms of Service more transparent and beneficial for fairer infrastructures.


Take part in the questionnaire until March 30th, 2022

Following the responses to the questionnaire, a final step will consist in a focus group to refine the selected prototype frameworks. Do not hesitate to contact us if you are interested (contact@apitos.org).

 

[1] The project offers a classification in different levels of ToS, going from green for fair services that respect users' rights and their data to red for those that raise serious concerns.

[2] For example, in 2013 and 2018, Twitter dealt a blow to its entire app ecosystem by abruptly changing the API terms of service. Although CEO Jack Dorsey apologized and said he wanted a better relationship with developers in 2015, the updates to these platforms destabilized existing ecosystems.

[3] Open access to scientific publications is limited today by rights restrictions on the use of APIs from scientific publishers who place a copyright on their API output (text and data mining).