7.2.1 Chatham House

British and American delegates to the Paris Peace Conference after World War I originally conceived of what would become Chatham House as a community of cross-Atlantic experts “to study international problems with a view to preventing future wars.” Then called the British Institute of Foreign Affairs, the think tank moved to Chatham House in 1923 and received its official Royal Charter in 1926. To this day, it remains a locus of expert discussions and global problem solving in foreign affairs, economic progress, and sustainability (Chatham House, History). Despite its participation in many of the most important worldwide debates in the last century, the community is primarily known for one thing: the Chatham House Rule.

The Chatham House Rule (“Rule”) is a simple rule of non-attribution used far beyond the confines of the Chatham House itself. The Rule reads as follows:

The Rule allows “people to speak as individuals, and to express views that may not be those of their organizations.” Therefore, the Rule “encourages free discussion” (Chatham House Rule FAQ) and knowledge production by establishing dissemination rules and norms of a given context, creating a sense of community around those rules, and threatening sanctions for violators. And the Rule can be modified in any given circumstance to fit the needs of a particular environment, thus providing the flexibility needed for knowledge production in different settings (Burnett Interview).

Keith Burnett, the Director of Communications and Publications at Chatham House, explains that the Rule sets the terms for information flow. Any event, meeting, or discussion governed by the Rule would either begin with an acknowledgment that the Rule applies – “This meeting is governed by Chatham House Rule” – or participants “would know in the invitation” to the event. Laying out the ground rule at the beginning has several important effects, according to Burnett. First, it lets meeting participants set their expectations for the shelf life, if any, of their statements during the event. Burnett noted, for example, that “whoever is running the meeting could take away some information and use it” to write a report or address some ongoing problem. Participants come into the meeting knowing this and can calibrate their participation accordingly. Second, the Rule encourages free flow of ideas by protecting participants from any negative effects of open sharing. The Rule is “there to help people share information that they learn in the meeting” without having to “worry about attribution” and any retribution that can result. Third, it prioritizes ideas and solutions over people and egos. Chatham House was established to solve important and vexing problems of international affairs, so sharing ideas is essential: “the idea is to free people up [to speak], free people up to feel that the ideas are more important than the individuals sharing them.” Finally, putting everyone on the same page with respect to information use immediately creates a sense of mutual trust among participants; there is, at least, an initial expectation that participants will behave according to the norms set out by the Rule (Reference WaldmanWaldman 2018; 52–54). Chatham House itself makes this connection between the Rule and trust, noting that “the rigorous implementation of the Rule is crucial to its effectiveness and for Chatham House’s reputation as a trusted venue for open and free dialogue” (Chatham House Rule FAQ).

The Chatham House Rule, then, reflects the contextual nature of information flow. Like the American sociologist Erving Reference GoffmanGoffman (1959), who explained that differences in social behavior and information disclosure often depend on the rules and expectations associated with different contexts, the Chatham House Rule creates contexts conducive to free discussion. Goffman suggested that social life can be divided into “front” and “back” stages, like a play, with different forms of social interaction happening in each. In front, like in the dining room of a fancy restaurant, servers interact with patrons formally, conversation is restricted to business, and behaviors are constrained by strict rules. In back, like in a kitchen out of view of paying customers, servers can let their hair down, curse, talk about their lives, and behave more informally among their colleagues. There are many front and back stages: work/home, professional/social interactions, when the kids are awake/after bedtime, and so forth. Each context maintains different rules and expectations, which permit or constrain social behavior accordingly. As Julie Reference CohenCohen (2000) has noted, this account of social behavior is not really about hiding from others; rather, it is about establishing the parameters of social space in ways that make continued interaction possible. That is precisely what the Chatham House Rule does, as well. It establishes the parameters of information flow in a meeting governed by the Rule, allowing participants to act more informally, engage more freely, and, hopefully, work more effectively to achieve the meeting’s goal. As Burnett notes, the Rule clarifies what is and is not permissible: “the whole purpose of the Rule,” Burnett notes, “is to share ideas without identifying the speaker, the source of the information, so if you do anything to identify that person, directly or indirectly, you should avoid it.”

Although one has to apply, anyone can join Chatham House in London (Chatham House Individual Membership). For Chatham House events and meetings held under the Chatham House Rule, it is the expressive power of the Rule that creates a sense of community among participants. A meeting governed by the Chatham House Rule immediately binds all participants together in a secret-keeping endeavor (Reference SimmelSimmel 1906). This creates a sense of trust among the members of the group that they are all present for the same purpose, on the same terms (Reference WaldmanWaldman 2018, 52–54). Although Chatham House is a “global community” and that “academics come from all over” to participate, Burnett suggests that the Rule, and the information flow it facilitates, is what matters. That is, it doesn’t matter who says what, “the whole point is to come and share knowledge – we share ours, you share yours,” and the Rule creates the space for that to happen (Burnett Interview). This is also part of the ethos of meetings run under the Rule, which allows participants new to Chatham House or the Rule to integrate into the community norms rather quickly.

The threat of sanctions also plays that role. Chatham House explains that although its Rule is “not legally binding, Chatham House will take disciplinary action against a member or guest who breaks the Rule,” which can mean “exclusion from all institute activities” (Chatham House Rule FAQ). Participants in Chatham House events governed by the Rule also mutually reinforce the Rule. The FAQs and Burnett both note that individuals ask other members if a given example of sharing – on social media or to other people – is permissible, with the overarching response focused on sharing without any kind of attribution. In this way, formal and informal sanction from the governing structure, as well as reinforcing work from other participants, work to ensure the Rule, and the meetings it governs, will achieve their goals.

This discussion suggests that the Chatham House Rule may be an effective form of privacy governance that fosters knowledge production within specific groups. By setting the parameters of information flow, it creates a sense of trust among participants, allowing them to rely on the confidentiality and discretion of a community built around the Rule.

7.2.2 Gordon Research Conferences

Like the Chatham House Rule, the Gordon Research Conference is a replicable institutional structure where privacy governance can foster knowledge production. The GRC was started in 1931, when the chemist Neil Gordon brought together scientists to network and share research results at Johns Hopkins University. Since then, the GRCs have grown worldwide, having attracted more than 30,000 participants annually to 300 conferences across the world (Gordon Research Conferences, History of GRC). They are well-regarded, invite-only, and attract what one participant called “the best people” (Sanfilippo Interview). Like Chatham House, GRC integrates a simple privacy rule into institutional governance; unlike Chatham House, its rule is strict confidentiality, not nonattribution. Like most knowledge commons, GRCs depend on other governance strategies beyond the privacy rule. For example, to maintain a close-knit community, each GRC is kept small and focused on a narrow topic at the frontier of the field. Secluded locations are preferred because seclusion is seen as something that helps structure the community, create a forum where people can get together and build relationships and trust, and avoid distractions.

The GRC No Publication Policy (“Policy”), sometimes referred to as the “off-the-record policy,” states, in relevant part: “To encourage open communication, each member of a conference agrees that any information presented at a Gordon Research Conference … is a private communication … and is presented with the restriction that such information is not for public use.” The Policy prohibits audio or video recording and photography. It also prohibits participants from preparing any form of publication based on the conference proceedings. The Policy covers any form of information sharing, including through social media, and covers formal presentations, poster sessions, and even informal discussions and conversations among participants. Yet the Policy is not an absolute prohibition; it vests authority in the individual making a presentation to consent to quotation, publication, recording, or other deviations from the rule: “[Y]ou [must] have the explicit written permission of the person you’re planning to quote. So, it’s allowable, but it’s kind of locked down” (Grannas Interview).

The primary purpose of the Policy is to give participants the confidence that they can share their latest, in-progress research without fear of others taking their ideas and publishing them first (Grannas Interview). Amanda Grannas, Associate Vice Provost for Research and Professor in the Department of Chemistry at Villanova University, explained how at the GRCs, she and other participants presented early stage research, data, and hypotheses without fear of being “scooped” by other participants. Her post-doc advisor had encouraged her to attend her first GRC to present work in progress to receive feedback and network with others. For Grannas, this experience stood in stark contrast to the American Chemical Society national meeting routinely attended by 12,000–15,000 people, where she would not present early work because the risk of being scooped was real. Grannas had been warned and knew of people whose ideas had been taken at the conference. During our interview, she praised GRCs for being a trusted community and forum and emphasized how important it was to be able to present pre-publication research in an environment where the rules facilitated free and constructive feedback on in-progress works.

Therefore, the GRC’s No Publication Policy contributes to knowledge production through a social process similar to that of the Chatham House Rule: a clearly defined rule sets participant expectations, building trust through community and the threat of sanctions. And those expectations are communicated to participants clearly and repeatedly up front. According to Grannas, acceptance letters to GRCs state “congratulations, … here are all kinds of expectations … . It’s expected that you’re not sharing this information, you’re not quoting any of this, you can’t refer to anything.” The Policy is also on the GRC webpage and conference chairs state the rules at the beginning of the event. “When the group comes together for the first time, there is a welcome from the chair and … they sa[y] explicitly that you can have a great time at Gordon” and even have a Gordon hashtag on Twitter, “but do not talk about hearing a great talk, … do not take a picture of a slide and tweet it or put it on Facebook.” There is, then, a real sense in which “what happens at Gordon, stays at Gordon” (Grannas Interview).

The confidentiality policy reinforces a sense of community by relying on participants to enforce it and, as noted earlier, by creating a network of secret-keepers (Reference SimmelSimmel 1906), thus creating the trust necessary for open discussion. The GRC Policy “facilitates community” by permitting open “coordination, collaboration, and direction of research” (Sanfilippo Interview). The Policy, in other words, solves a problem particularly salient to academic researchers: “how to facilitate discussion without worrying about people stealing your ideas” (Id.).

When veterans see new participants, “the chain keeps going and norms and rules are passed on” (Grannas Interview). Sometimes this happens explicitly. One participant recalled hearing presenters say, for example, “In the spirit of the GRC, I’m showing you unpublished research,” which implies a duty to maintain the speaker’s confidentiality (Sanfilippo Interview). Otherwise, participants “learn [the norms] just by observing the practice” (Id.). Participants recognize that GRCs include a relatively “small number of people” where “everybody knows everybody, except for new people coming in and we want to do right by them” (Grannas Interview). GRCs include social events for networking purposes, including group meals, mentorship opportunities, and other “non-sciencey” meet-ups (Id.). These not only promote additional knowledge production but also serve to pass on appropriate expectations of social behavior to the next generation of scholars. Invited attendees are encouraged to bring guests who can participate in daily social activities, but to preserve confidentiality, “guests are not permitted to attend the conference lectures, or poster sessions.” (Gordon Research Conferences, Policies; Grannas Interview). As Grannas noted, the GRC has become “more family friendly. They very much lock it down that guests cannot attend the science. Only the scientists are there. And I think that is kind of leading towards the privacy issue. It’s easy for someone that isn’t as familiar with the expectations to start tweeting about, oh this cool thing that I’m sitting at with my wife or my husband at this thing.”

Also like Chatham House, GRCs mostly rely on participants to enforce privacy governance. Participants who notice someone flouting the Policy may “contact the chair and say, ‘look, this happened, I don’t know what you want to do, but maybe this person shouldn’t come back to one of these” (Id.). More commonly, though, the pressure to adhere to GRC confidentiality norms is social and informal. According to Grannas, there is a “level of accountability that, I think, the small nature of the group [provides]. … [T]he fact that you’re going into it with very explicit goals, it would be very antithetical to then take an idea and go off and scoop or do whatever nefarious thing you’re going to do.” Violators are “blackballed” in the scientific community if they record, tweet, or scoop. “We would take care of it,” Grannas noted, if it actually happened.

Notably, the GRC Policy is different from the Chatham House Rule. The former is a confidentiality requirement, whereas the latter permits disclosure of the substance of discussions just without names and attribution. Although both are privacy governance structures, they serve both overlapping and distinct purposes. The goal of the Chatham House Rule was, from the beginning, to foster open conversation to solve a pressing problem in world affairs. The discussion would inform the meeting’s chair, who would likely take what she had learned and incorporate it into public facing reports, recommendations to policymakers, or proposals for peace. The GRC No Publication Policy also fosters conversation and sharing ideas by creating the space necessary for researchers to trust they won’t be scooped. Indeed, its expansion over the last few decades is a testament to its success at creating that kind of environment. But the goal of the GRCs is to help participants, not the outside world, learn and receive feedback on their own research projects. As such, different rules regarding dissemination make sense. Both rules, however, leverage the same social process by which privacy governance fosters knowledge production.

7.2.3 Broadband Internet Technical Advisory Group

The Broadband Internet Technical Advisory Group (BITAG) was created to bring together technical experts on a voluntary basis to discuss and develop consensus advisory reports on technical operation of the Internet. Here, there is no risk of being scooped, as there is at GRCs. Like the original Chatham House meetings, BITAG members come together to produce a public-facing document. But the need for privacy governance to create open and frank conversation remains; stakeholders and participants come from different sectors, business, and backgrounds, and they are trying to solve potentially contentious technical problems. To do that, members must be free to share inchoate or developing ideas (Reference RichardsRichards 2015) and offer suggestions that their employers might oppose. To achieve this goal, BITAG sets out a strict confidentiality rule that leverages the same social process of fostering knowledge production through trust, community, and the threat of sanctions.

The BITAG restricts the dissemination of “all information disclosed by any Participant during any meeting or activity” of the BITAG Technical Working Group or committees (“BITAG information”).Footnote ⁵ All such information is considered confidential, and participants agree that they “shall not use any portion of the … information for any purpose except to perform his, her, or its obligation to BITAG” (BITAG Intellectual Property Rights Policy (“IPR Policy”) Section 4.1). BITAG also binds itself to not use or disclose confidential information shared during BITAG meetings and working group sessions (Id.). There is some leeway to share information with other BITAG members. Membership in BITAG extends beyond the individuals who participate in meetings to the companies and organizations of which individuals are a part. Thus, for example, if one company’s employee attends a BITAG meeting, that employee may disclose BITAG information (but not “highly confidential information”Footnote ⁶) to other “employees, contractors and agents of” her company, but only if they “agree to maintain the confidentiality of the BITAG Information” (BITAG IPR Policy Section 4).

The BITAG confidentiality rule “helps, especially when there is disagreement” to ensure open and frank discussion. Jason Livingood, Vice President of Technology Policy and Standards at Comcast and a BITAG participant, notes that the point of the rule is to “learn and move forward” together and “talk about things openly.” To do that, participants need to know that “there’s not going to be any attribution; we’re not going to go back to our bosses and say, ‘this other guy at this other company said’” something at a BITAG meeting. What’s more, BITAG wants to encourage expert participants to speak openly, even if what they say conflicts with their employers’ official position. Therefore, the expectation is that conversations “are very private and in particular may surface something about my company’s perspective, but I personally disagree with. … If someone were to share that, I would be in a lot of hot water professionally” (Id.). In this way, the rule fosters sharing and knowledge production by protecting participants from professional retribution.

To ensure that all participants can operate under the same set of confidentiality expectations, BITAG deploys the same strategy as Chatham House and GRC: event chairs state the rule at the beginning and participants reinforce it over time. At the start of working group sessions, for example, leaders will “go back for the standards … [and say,] ‘here, just as a reminder, this is how we work, we are not going to tweeting about what we’re talking about here to sharing anything we talk about.’ Everything is confidential” (Livingood Interview). Throughout the process of working together to achieve some kind of consensus report, “there are reminders, [like,] ‘hey, don’t share any of this externally. There will be a point in time when we’re ready to do that, but it’s not right now’” (Id.). This set clear expectations and fosters trust among BITAG working group participants by setting the terms of social interaction and expressing what the organization feel is proper behavior. If participants found out that someone was disclosing the substance of their discussions, members “would be very upset.” It would, Livingood noted, “be a real violation of trust.”

According to Kaleb Sieh, the Deputy Director and General Counsel of BITAG, Inc., the confidentiality and nondisclosure rules create a community of members: “confidentiality binds the entire organization” because everyone knows that “we only share with people as necessary.” BITAG further develops this community sentiment by design, creating working groups and committees that have professional membership requirements. This allows participants to trust that they’re getting the best input and advice on their projects (Livingood Interview). Membership rules aim to (i) maintain openness to “both individuals and organizations representing stakeholders from a broad cross-section of the Internet ecosystem” (Reference Sieh and HatfieldSeih and Hatfield 2012, 8) while (ii) ensuring technical expertise among the working group participants and (iii) guarding against “forum packing” or other strategic behavior by particular stakeholders. In other words, membership and some associated voting rules aim to maintain the integrity of the community and trust within and from outside the community.

This community takes it upon itself to police and enforce the confidentiality rules, even though there has been little need for enforcement (Livingood Interview). Although individuals and companies can be terminated from BITAG for violating the rules, the most important sanctions are more organic and social: “if they violate the rules, their professional reputation would be shot” (Id.). Individual participants, the scientific experts and engineers, often know (or know of) each other from the broader scientific and technological community or even from other similar working groups or standard setting organizations, such as the Internet Engineering Task Force. While governance rules vary across organizations and contexts, the professional norms and reputations span them (Id.). More importantly, violations of BITAG confidentiality represent breaches in trust (Reference WaldmanWaldman 2018). As Livingood noted, those who disseminate information are “definitely no longer part of the group. … But even other members would make it clear that you can’t trust” those who violate the rules. “Not trusting on an individual level is important,” and when BITAG members can’t trust someone, they are ostracized from the community, even if they’re not formally expelled.