Dr. Christoph Riedl provided mentorship, feedback, and methodological assistance in developing this study. The full text of this paper is available on the APSA Preprint Server. This content is an early research output and has not been peer-reviewed at the time of posting.

The extent to which human rights considerations are integrated into Internet technical standards shapes how people exercise their rights online. Understanding the policy implications of engineering decisions helps support the Internet as a right-enabling environment. Prior literature has provided valuable evidence on how values and rights are (or are not) considered in technical processes and organizations. This study contributes to our understanding of human rights in Internet governance by exploring the Internet Engineering Task Force (IETF) and Internet Research Task Force (IRTF) email network at scale. We analyze all emails sent by IETF and IRTF participants recorded in the public IETF Mail Archive, containing the term “human rights”, sent from 1992 to early 2025 (n = 3,377). Emails are annotated by a large language model to indicate the author’s position on integrating human rights. They are then analyzed using time series, regression, and network analysis. This analysis reveals how processes of norm entrepreneurship regarding human rights considerations in Internet standards unfolded over the course of 33 years. We find evidence of a new cohort of norm entrepreneurs emerging around 2013. The timing of this new cohort aligns with increasing interest in human rights online from international organizations and many of the participants write from emails ending in org, implying an affiliation with organizations. They are also more supportive of human rights, on average, than the prior cohort and several of them send a very high number of emails, potentially indicating they are acting as norm entrepreneurs to promote human rights. Based on our network analysis, it appears that these entrepreneurs had limited success in diffusing norms to the broader community. This study provides insight into how processes of norm entrepreneurship and diffusion evolved in the IETF and IRTF community. It builds on prior literature using email data to map normative evolution with a novel email dataset. Understanding these informal Internet-mediated interactions can help explain policy-making outcomes in Internet governance. 

Introduction

The integration of human rights into Internet infrastructure is controversial. Resistance to integrating human rights primarily takes the form of conflict over norms around the characteristics of the Internet and the role of engineers responsible for developing and maintaining it.  Following Finnemore and Sikkink (1998, p. 891), we consider norms “a standard of appropriate behavior for actors with a given identity…”. According to Radu et al. (2021, p. 3), “Entrepreneurs – be it state or non-state actors – define and establish standards of behaviour that become influential within the community mostly by adhering to relevant networks”. These norms then diffuse further, in what Finnemore and Sikkink (1998, p. 895) have termed a ‘norm cascade’. Many in the IETF believe that engineers should prioritize maintaining the neutrality of technology and the ability of any device to join the network and connect with other devices (Cath and Floridi, 2016; ten Oever, 2023) over other objectives. Human rights norm entrepreneurs in technical communities have faced entrenched resistance, not necessarily due to disputes over the importance of human rights in general, but due to concerns that integrating human rights considerations into technical work would jeopardize these deeply held values. For example, one IETF engineer interviewed by Cath (2021, p. 11) stated that “All that [standards] needs to be consistent and predictable, you cannot create a suite of human rights considerations for protocols, then a completely different suite for whatever because they need to be interoperable. Everything on the Internet needs to be interoperable in order to make sense.” In this instance, integrating human rights is seen as potentially jeopardizing interoperability (Cath, 2021), leading to resistance to integrating human rights considerations. We analyzed the online processes surrounding the integration of human rights considerations into technical standards. We applied time series, regression, and network analysis to a novel dataset of IETF and IRTF emails pertaining to integrating human rights considerations into technical standards. This analysis reveals how processes of norm entrepreneurship and diffusion evolved in the Internet Engineering Task Force and Internet Research Task Force community.

The Internet Engineering Task Force (IETF) is a standards development organization (SDO) for the Internet. SDOs create and refine the standards that allow different entities on the Internet to connect; developing and maintaining the Internet as a communications network. According to the IETF’s mission statement: “the overall goal of the IETF is to make the Internet work better” (Internet Engineering Task Force a, n.d.). Cath and Floridi (2017, p. 2) write that the IETF is “one of the most influential SDOs, capable of making the most significant difference in the architectural implementation of values or human rights.” While the ability of SDOs in general to influence human rights is questioned by some scholars (Mueller and Badiei, 2019), the IETF is certainly an important player in this space. Individuals voluntarily participate in the IETF in their personal capacity. According to its website, “The IETF welcomes all interested individuals and participants come from all over the world and from many different parts of the Internet industry” (Internet Engineering Task Force a, n.d.). According to Cath and Floridi (2017), at the time of their study, IETF participants were mostly white, male, and Western and worked for large companies. Participants develop technical standards through an open process of offline (Andrade et al, 2018) and online collaboration. This online collaboration often takes the form of email exchanges on public mailing lists. 

The Internet Research Task Force (IRTF) works in parallel with the IETF on “longer term research issues related to the Internet” (Internet Research Task Force, n.d.). It is composed of 16 research groups focused on Internet technology, applications, architecture, and protocol-related topics. The Human Rights Protocol Considerations Research Group (HRPC) is particularly relevant to this study. It researches “whether standards and protocols can enable, strengthen or threaten human rights, as defined in the Universal Declaration of Human Rights (UDHR) and the International Covenant on Civil and Political Rights (ICCPR), specifically, but not limited to the right to freedom of expression and the right to freedom of assembly” (Internet Research Task Force b, n.d.) While IETF participants tend to come from industry, many IRTF participants are members of the academic research community or non-governmental organizations. 

As the Internet becomes more central to daily life, the impact of online harms and the importance of protecting the Internet as a rights-enabling space have become more pronounced. This has been recognized by international organizations and is increasingly acknowledged by the technical community. In 2011, a United Nations report highlighted the role of the Internet in enabling human rights (UN Human Rights Council, 2011). One year later, the UN Human Rights Council published the resolution on “The promotion, protection and enjoyment of human rights on the Internet” (UN Human Rights Council, 2012). Also in 2012, the Internet Society published the report “Human Rights and Internet Protocols: Comparing Processes and Principles” (Liddicoat and Doria, 2012). In 2013, the Snowden leaks were published, revealing U.S. government surveillance (British Broadcasting Corporation, 2013). These events were followed by more formal recognition in Internet governance technical organizations, including the IRTF and Internet Corporation for Assigned Names and Numbers (ICANN), of the relevance of human rights to their work. In 2015, the IRTF began to formally consider and study human rights by chartering the Human Rights Protocol Considerations Research Group (Internet Research Task Force b, n.d.). This group was chartered one year after the idea to “research how Internet standards and protocols impacted human rights values and develop guidelines that IETF engineers could use to mitigate the negative impact of their work” was presented to the IETF (Cath, 2021, p. 2-3). A 2016 update to ICANN bylaws also included human rights for the first time (Internet Corporation for Assigned Names and Numbers, n.d.). The work of the HRPC resulted in the publication of Request For Comments (RFC) 8280 in 2017 (Cath and ten Oever, 2017) and RFC 9620 in 2024 (Grover and ten Oever, 2024), both of which aid in the integration of human rights considerations into technical standards. As this study shows, these advances emerged from processes of norm entrepreneurship and diffusion.

Methods

To map norm development over time, this study integrates time series, regression, and network analysis. Emails are annotated by OpenAI’s GPT-4o, a large language model, to designate whether they support, oppose, or are neutral towards integrating human rights considerations in technical standards (Open AI Platform, n.d.). Methodological and prompt details are available in the full paper on APSA Preprints. Emails are assigned a score of 1 if they support integrating human rights, -1 if they oppose integrating human rights, and 0 if they are neutral or the email is not relevant. The prompt provided to GPT included criteria for each score, as well as examples. Email data is then graphed in a time series to reveal overall changes in activity level and average support. Sending email accounts are also categorized by top level domain (e.g. com or edu), providing insight into the affiliation of the sender. Top level domain (TLD) and how many emails an account sent (network out-degree) were then regressed on support. These results were used to group email accounts into cohorts based on when they joined the network. Cohorts are then compared on the most frequent top level domain and the average level of support over time. Lastly, cohort networks are compared. 

Results

Our results suggest the presence of two distinct phases of norm entrepreneurship in the network. We use these phases to separate participants into two cohorts based on when they joined the network. This is similar to the method used by Milan and ten Oever (2017) to analyse cohorts in the Internet Corporation for Assigned Names and Numbers. We then evaluate the network, participant attributes, and level of support for each cohort to explore norm entrepreneurship and diffusion. 

Figure 1 displays the number of emails per year for the top five most frequent top level domains. The top level domain provides insight into the potential affiliation of the sender. For example, an email ending in edu indicates that the sender is likely affiliated with an academic institution, while an email ending in com is likely a personal or company email.

Figure 1 shows that for most years com is the most prevalent top level domain, with the exception of the years 2015, 2016, 2017, 2020, and 2025. In 2025, edu was the most prevalent TLD (though this is likely due to the limited data for 2025). In the other years, org was either the most prevalent TLD or tied with com (as in the year 2020). This could indicate that participants associated with organizations, potentially members of civil society, were the most active participants on the network in 2015, 2016, and 2017. Trithara (2024, p. 2), citing Abbott and Snidal (2009, p. 64-5), summarizes civil society as including “a diverse range of value- and interest-driven actors that seek to shape the rules of social life.” In this context, one example of a civil society organization is Access Now, which promotes digital rights. 

Levels of support for integrating human rights considerations varied by TLD. Table 1 displays the total average support score for each TLD for all years as well as the number of total emails sent from that domain. Average support score is the sum of all support scores for emails ending with that TLD divided by the total number of emails sent by the TLD. 

Email Top Level Domain	Average Support Score	Total Number of Emails Sent
edu	0.44	108
com	0.22	1417
net	0.47	372
org	0.62	937
fr	0.46	101

Table 1. Top Level Domain Average Support and Frequency

These average scores indicate that email authors writing from addresses ending in com displayed the lowest average support for integrating human rights considerations. Authors writing from addresses ending org, likely organizations, displayed the highest average support. Figure 2 shows the average support score per year by TLD.

For most years, emails ending in net, org, or edu showed the highest level of support, while emails ending in com showed the lowest level of support. When comparing average scores for emails ending in org across years, there is a marked increase in average scores in 2013 and after compared to prior years.

Figure 3 displays the network graph of all emails in the dataset. Nodes are either sender or receiver email accounts. They are sized based on the number of emails sent from that account. They are colored based on the account top level domain: addresses ending in com are pink, addresses ending edu are orange, net is green, fr is yellow, and org is purple. Accounts associated with mailing lists are colored black. Each edge represents an email sent between two nodes. Since one email can go to multiple recipients, each edge represents one sender-receiver dyad. For example, an email from one account to three others would be represented in three separate edges. Edges are colored based on the support score of the email: -1 is red, 0 is grey, and 1 is blue. Due to privacy considerations, email accounts associated with individuals are not labeled with their addresses. Mailing list email accounts that received 10 or more emails are labeled with their addresses. For 177 emails (5.2% of the dataset), the mailing list address could not be automatically extracted. In total, this network has 4,998 edges (email dyads) and 971 nodes (email accounts). 

The most prominent email lists in this network are the Human Rights Protocol Considerations Research Group mailing list (hrpc@irtf.org), used for the work of the HRPC research group, and the IETF-Discussion mailing list (ietf@ietf.org), a general discussion list. The average support score for emails sent to the hrpc@irtf.org mailing list is 0.51, while the average support score for emails sent to the ietf@ietf.org mailing list is 0.14. These differences are also visually reflected in the network, with noticeably more blue lines connected to the hrpc@irtf.org node than the ietf@ietf.org node. 

Regression

This section explores the extent to which affiliation and email activity are associated with support. It addresses the questions: are email authors associated with organizations or universities more supportive of human rights than authors with other affiliations? And are authors who send more emails more supportive of human rights than authors who send fewer emails? Top level domain (TLD), a proxy for affiliation, and out-degree (the number of emails sent from an account) were regressed on support using OLS regression from the statsmodels module in Python. In this data structure for regressing domain on support the rows are individual emails and the columns are attributes of each email, including its support score and the TLD of the sender email address. No top level domain was a significant predictor of support, indicating that affiliation does not significantly impact support. 

Regressing out-degree on support was done first for all years then for individual years to determine change over time in the relationship between out-degree and support. Out-degree was calculated for each email address based on how many emails it had sent for all years to how many other email addresses. For example, a sender address sending two emails, each with two recipients, would have an out-degree of four.  In the data structure for this regression, the rows were unique email addresses and the columns were their out-degree in the network and the average support across emails they sent. For example, if a sender address sent one email that was supportive (1) and one email that was neutral (0), they would have an average support of 0.5. Out-degree was then regressed on average support. The results show that out-degree is a significant predictor of average support (p < .01) though the effect is very small: the coefficient is 0.002. This calculation was then repeated for each year in the dataset. There were no individual years in which the relationship was significant. 

The relationship between out-degree and average support was then assessed for the time period before 2013 and the time period during and after 2013. The year 2013 was selected due to the marked increase in activity in the years following and the increase in average support for integrating human rights.  In the network of emails sent between 1992 and 2012, out-degree was not a significant predictor of average score. However, in the network of emails sent between 2013 and 2025, out-degree was a significant predictor of average score (p < .01), though the impact is still small (coefficient is 0.0032). This indicates that there was a shift in communications patterns among participants who joined before 2013 and those who joined during and after 2013. For those who joined before, the number of emails they sent is not related to their level of support. For those who joined during and after 2013, the number of emails they sent is related to their level of support. This could indicate a deliberate effort by norm entrepreneurs to spread support for human rights by sending many emails. 

Cohort Comparison

This section examines the question: how have network participation and communication patterns evolved over time? Based on the regression results and email activity, a new cohort of norm entrepreneurs sending many emails supporting the integration of human rights considerations may have joined the network around 2013. Email accounts are grouped into cohorts based on whether they joined the network before 2013 (cohort 1) or during or after 2013 (cohort 2). Figure 4 shows the yearly number of emails sent by each cohort. 

Cohorts were compared by the most frequent top level domain, the average level of support over time, and their network characteristics. 

Top Level Domain

The total number of emails sent for each of the ten most frequent domains and their average level of support for cohorts 1 and 2 are displayed in tables 2 and 3. 

Email Top Level Domain	Total Number of Emails Sent	Average Support
com	275	0.05
net	73	0.26
fr	63	0.49
org	46	0.07
es	14	0.07
tv	8	0.13
edu	8	0.13
jp	7	-0.43
im	6	0.50
uk	4	0.25

Table 2. Most Frequent Domains in Cohort 1. For 30 emails in this cohort, the top level domain could not be automatically extracted. 

Email Top Level Domain	Total Number of Emails Sent	Average Support
com	1142	0.26
org	891	0.65
net	299	0.52
edu	100	0.46
ie	73	0.18
fr	38	0.39
it	28	0.29
uk	27	0.22
ch	25	0.48
ca	21	0.52

Table 3. Most Frequent Domains in Cohort 2

For cohort 2, the prevalence of emails ending in org and edu is much higher than for cohort 1. These domains also have higher average support scores in cohort 2 than in cohort 1. Emails ending in com also saw an increase in average support score from cohort 1 to cohort 2, though the change was not as large as for org and edu.  

Average Level of Support

Figure 5 shows the yearly average support for each cohort. 

With the exception of 2014, the average support of cohort 2 is equal to or greater than the average support of cohort 1 for each year after 2013. 

Figure 6 shows the full email network with nodes colored by cohort. Nodes belonging to cohort 1 are colored orange and nodes belonging to cohort 2 are colored aqua. 

This network shows that many of the high out-degree cohort 2 nodes (email accounts that send many emails) are displayed near the hrpc@irft.org mailing list, which is receiving many supportive emails. These nodes can be considered norm entrepreneurs. The nodes displayed near ietf@ietf.org are more mixed between the two cohorts and generally have a lower degree. Emails sent to ietf@ietf.org are also more distributed in terms of supportive and opposing emails. The email list hrpc@irtf.org is more frequently emailed by cohort 2 accounts than by cohort 1 accounts, and these emails appear to be mostly supportive. Across the two cohorts, these email lists receive emails with differing levels of support. Emails sent by cohort 1 accounts to ietf@ietf.org have an average support of 0.18, while emails sent by cohort 2 accounts have an average support of 0.13. Emails sent by cohort 1 accounts to hrpc@irtf.org have an average support of 0.16, while emails sent by cohort 2 accounts have an average support of 0.54. For 177 emails, the mailing list address could not be automatically extracted. 

Given the increasing interest of international organizations in human rights online starting around 2011, and the affiliation of cohort 2 participants with organizations and academia, and their increased participation around 2013, it seems likely that many of these participants are members of civil society who joined to help carry out the priorities of international organizations by promoting human rights. This affiliation would provide these participants with a different background and set of priorities than the participants affiliated with large companies who have traditionally conducted most of the work of the IETF. The association of many cohort 2 participants with international organizations would explain their increased activity around the same time organizations began taking note of human rights online and their higher levels of support for human rights. 

Discussion

As discussed above, the years 2014 to 2016 saw considerations of human rights become increasingly formalized in the Internet governance community. These policy developments are reflected in the analysis conducted in this study. The entry of the second cohort of participants, many of whom are affiliated with organizations, into the email network around 2013 aligns with the increasing interest of international organizations in human rights online. This increasing interest is also reflected in the increase in yearly average support scores around 2013. Both of these developments predate, and potentially help explain, the chartering of the HRPC in 2015. 

Much of the prior literature analyzing email data from the Internet governance community focuses on particular thematically-relevant email lists (Becker, ten Oever, and Nanni, 2022; Doty, 2015), with the notable exception of Welzl et al (2021). Our study shows how support for integrating human rights varies across email lists, particularly email lists that are not officially delineated as relevant to human rights. If only emails on the list hrpc@irtf.org had been included in the analysis, many of those who are critical of integrating human rights considerations would have been left out. This would have created a false picture of homogeneity and consensus in the email network, while examining all email lists where the term “human rights” is used reveals the contention and variety of viewpoints in the debate. 

This study also furthers our understanding of norm entrepreneurship and diffusion in Internet governance. It shows that a second cohort of network participants became active on the network around 2013. Compared to the first cohort, who joined the network between 1992 and 2012, the second cohort had more participants affiliated with organizations. The second cohort potentially included human rights norm entrepreneurs, distinguished by their relatively high number of supportive emails. Despite their level of activity, these norms may not have diffused to the rest of the network. Time series analysis of average levels of support among those who joined before 2013 reveals that their average level of support was consistently lower than those in the newer cohort. Network analysis reveals that the activity of potential norm entrepreneurs was primarily concentrated around the email list hrpc@irtf.org. The average level of support of emails sent to the list ietf@ietf.org, a more general purpose discussion list, declined from the first cohort to the second cohort. These results indicate that while norm entrepreneurs may have been active on the network, their levels of support for integrating human rights considerations in technical standards may not have diffused to the prior cohort of participants or to more general discussion in the IETF. 

Conclusion

This study shows that the dynamics of norm development can be quantitatively measured by integrating annotation by a large language model with time series, regression, and network analysis; allowing for a more comprehensive overview than would be possible with exclusively qualitative tools. It provides insight into the development and diffusion of norms pertaining to integrating human rights considerations into technical standards on IETF and IRTF email networks.  This furthers our understanding of how informal Internet-mediated interactions shape policy-making outcomes in Internet governance. Our results show that while norm entrepreneurs were able to shape discourse on one part of the network (the hrpc@irtf.org mailing list), they had less success in diffusing norms to other parts of the network (e.g. the ietf@ietf.org mailing list). This shows the importance of collaboration across divides for integrating human rights into Internet infrastructure. 

References

Abbott, K. W., & Snidal, D. (2021). The governance triangle: Regulatory standards institutions and the shadow of the state. In The spectrum of international institutions (pp. 52-91). Routledge.

Andrade, L., Braga, J., Pereira, S., Roque, R., & Santos, M. (2018). In-Person and Remote Participation Review at IETF: Collaborating Without Borders. arXiv. https://arxiv.org/abs/1805.09140 

Becker, C., ten Oever, N., & Nanni, R. (2022). The Standardisation of Lawful Interception Technologies in the 3GPP: Interrogating 5G and Surveillance Amid US-China Competition. SSRN. http://dx.doi.org/10.2139/ssrn.4167105

British Broadcasting Corporation. (2013, August 20). Edward Snowden: Timeline. https://www.bbc.com/news/world-us-canada-23768248

Benthall, Sebastian. 2015. “Testing Generative Models of Online Collaboration with BigBang.” In , 182–89. https://conference.scipy.org/proceedings/scipy2015/sebastian_benthall.html

Cath, C., & Floridi, L. (2017). The Design of the Internet’s Architecture by the Internet Engineering Task Force (IETF) and Human Rights. Sci Eng Ethics, 23(2), 449-468. https://doi.org/10.1007/s11948-016-9793-y 

Cath, C., & ten Oever, N. (2017). Research into Human Rights Protocol Considerations (Request for Comments: 8280). Internet Research Task Force. https://datatracker.ietf.org/doc/rfc8280/ 

Cath, C. (2021). The technology we choose to create: Human rights advocacy in the Internet Engineering Task Force. Telecommunications Policy, 45(6), 102144.

Cinelli, M., De Francisci Morales, G., Galeazzi, A., Quattrociocchi, W., & Starnini, M. (2021). The echo chamber effect on social media. Proceedings of the National Academy of Sciences, 118(9), e2023301118.

De Francisci Morales, G., Monti, C., & Starnini, M. (2021). No echo in the chambers of political interactions on Reddit. Sci Rep, 11(1), 2818. https://doi.org/10.1038/s41598-021-81531-x 

Doty, Nick. 2015. “Reviewing for Privacy in Internet and Web Standard-Setting.” In Security and Privacy Workshops (SPW), 2015 IEEE, 185–192. IEEE. https://ieeexplore.ieee.org/document/7163224/

Finnemore, M., & Sikkink, K. (1998). International Norm Dynamics and Political Change. International Organization, 52(4), 887–917.

Garimella, K., Morales, G. D. F., Gionis, A., & Mathioudakis, M. (2018). Quantifying controversy on social media. ACM Transactions on Social Computing, 1(1), 1-27.

Grover, G. & ten Oever, N. (2024). Guidelines for Human Rights Protocol and Architecture Considerations (Request for Comments: 9620). Internet Research Task Force. https://datatracker.ietf.org/doc/rfc9620/

Internet Engineering Task Force a. (n.d.). Introduction to the IETF. Retrieved April 7, 2025, from https://www.ietf.org/about/introduction/

Internet Engineering Task Force b. (n.d.). Mail Archive. Retrieved April 7, 2025, from https://mailarchive.ietf.org/arch/

Internet Engineering Task Force c. (n.d.). IETF Datatracker. Retrieved April 7, 2025, from https://datatracker.ietf.org/

Internet Research Task Force a. (n.d.). Overview. Retrieved April 7, 2025, from https://www.irtf.org/

Internet Research Task Force b. (n.d.). Human Rights Protocol Considerations Research Group HRPC. Retrieved April 7, 2025, from https://www.irtf.org/hrpc.html

Internet Corporation for Assigned Names and Numbers. (n.d.). Bylaws FOR INTERNET CORPORATION FOR ASSIGNED NAMES AND NUMBERS | A California Nonprofit Public-Benefit Corporation – ICANN. Retrieved April 7, 2025, from https://www.icann.org/en/governance/documents/bylaws-for-internet-corporation-for-assigned-names-and-numbers-a-california-nonprofit-public-benefit-corporation-icann-01-10-2016-en

Khare, P., Karan, M., McQuistin, S., Perkins, C., Tyson, G., Purver, M., Healey, P., & Castro, I. (2022). The Web We Weave: Untangling the Social Graph of the IETF. Proceedings of the International AAAI Conference on Web and Social Media, 16(1). 500-511. https://doi.org/10.1609/icwsm.v16i1.19310

Krishnamurthy, V. (2019). Are Internet Protocols the New Human Rights Protocols? Understanding ‘RFC 8280 – Research into Human Rights Protocol Considerations.’ Business and Human Rights Journal, 4(1), 163–169. doi:10.1017/bhj.2018.30

Liddicoat, J., & Doria, A. (2012). Human Rights and Internet Protocols: Comparing Processes and Principles. Internet Society. https://www.internetsociety.org/resources/doc/2012/human-rights-and-internet-protocols-comparing-processes-and-principles/

Marino, M. C. (2020). Critical code studies. MIT Press.

Milan, Stefania, and Niels ten Oever. 2017. “Coding and Encoding Rights in Internet Infrastructure.” Internet Policy Review 6 (1)

Mueller, M. L., & Badiei, F. (2019). Requiem for a Dream: On Advancing Human Rights via Internet Architecture. Policy & Internet, 11(1), 61-83. https://doi.org/10.1002/poi3.190 

Nanni, Riccardo. “Digital Sovereignty and Internet Standards: Normative Implications of Public-Private Relations among Chinese Stakeholders in the Internet Engineering Task Force.” Information, Communication & Society 0, no. 0 (October 1, 2022): 1–21. https://doi.org/10.1080/1369118X.2022.2129270.

Open AI Platform. (n.d.). GPT-4o. Retrieved April 7, 2025, from https://platform.openai.com/docs/models/gpt-4o 

Radu, R., Kettemann, M. C., Meyer, T., & Shahin, J. (2021). Normfare: Norm entrepreneurship in internet governance. Telecommunications Policy, 45(6). https://doi.org/10.1016/j.telpol.2021.102148

Riedl, C., Grad, T., & Lettl, C. (2024). Competition and Collaboration in Crowdsourcing Communities: What Happens When Peers Evaluate Each Other? Organization Science, 35, 1957–1976. https://doi.org/10.1287/orsc.2021.15163 

ten Oever, N., Milan, S., & Beraldo, D. (2020). Studying Discourse in Internet Governance through Mailing-list Analysis. In D. L. Cogburn, L. DeNardis, N. S. Levinson, & F. Musiani (Eds.), Research Methods in Internet Governance. Cambridge, MA: MIT Press. https://direct.mit.edu/books/oa-monograph/4936/chapter/625914/Studying-Discourse-in-Internet-Governance-through

ten Oever, N. (2023). Norm conflict in the governance of transnational and distributed infrastructures: the case of Internet routing. Globalizations, 20(1), 184-200.

Trithara, D. (2024). Agents of platform governance: Analyzing U.S. civil society’s role in contesting online content moderation. Telecommunications Policy, 48(4). https://doi.org/10.1016/j.telpol.2024.102735

UN Human Rights Council. (2011). Report of the Special Rapporteur on the promotion and protection of the right to freedom of opinion and expression, Frank La Rue (A/HRC/17/27). United Nations General Assembly. https://www2.ohchr.org/english/bodies/hrcouncil/docs/17session/A.HRC.17.27_en.pdf

UN Human Rights Council. (2012). The promotion, protection and enjoyment of human rights on the Internet : resolution / adopted by the Human Rights Council (A/HRC/RES/20/8). United Nations General Assembly. https://digitallibrary.un.org/record/731540?ln=en&v=pdf 

Welzl, M., Oepen, S., Jaskula, C., Griwodz, C., & Islam, S. (2021). Collaboration in the IETF: An Initial Analysis of Two Decades in Email Discussions. ACM SIGCOMM Computer Communication Review, 51(3), 29–32. https://dl.acm.org/doi/pdf/10.1145/3477482.3477488