At its technical plenary session at IETF 98 in Chicago, the Internet Architecture Board sponsored a lively debate about how best to handle human rights considerations in protocol development.
The discussion revolved around a draft document that was developed during the last two years by the Internet Research Task Force’s Human Rights Protocol Consideration (HRPC) Research Group. The HRPC group is exploring how Internet protocol development can enable, strengthen, or weaken rights, such as freedom of expression and freedom of assembly as outlined in widely approved treaties such as the Universal Declaration of Human Rights and the International Covenant on Civil and Political Rights.
The HRPC group believes that as a global network-of-networks, the Internet should strive to provide continuous connectivity to all users for all content. As such, the group believes that the Internet’s promise of open, secure, and reliable connectivity makes it a key enabler of human rights. The group is exploring the relationship between human rights and protocols, and is working on guidelines for protocol developers to help them avoid situations where a new protocol would inhibit users’ ability to exercise their freedom. Ideally, these guidelines will be similar to the work done for Privacy Considerations in RFC 6973.
Niels ten Oever, cochair of the HRPC Research Group and head of digital for article 19, opened the discussion by conceding that the HRPC group’s effort to understand and demonstrate the human rights impact of Internet protocols is difficult.
“We’ve all seen the Internet as this huge engine for freedom of expression, and it has indeed enabled us to create whole new opportunities for people to express themselves and gather information. But it doesn’t mean there aren’t downsides to our work as in the case of pervasive surveillance,” ten Oever explained.
Ten Oever pointed out that the Internet is playing an increasingly important role in such areas as freedom of expression, association, and assembly, as well as in education, public debate, and even voting. He emphasized that despite the IETF’s efforts, access to the Internet is not equally distributed to the rest of the world.
“We have, in the IETF, contributed greatly to the shaping of the Internet that we have today,” ten Oever said. “But with great power, comes great responsibility. This is a call to assert that power.”
Ten Oever shared that the IETF holds and propagates certain values, such as fairness, decentralization of control, and sharing of resources. So considering ways to mitigate how its protocols might be used to limit human rights is not out of the standards body’s charter. In particular, he pointed to RFC 6973, which outlines privacy considerations for Internet protocols, and to BCP 72, which provides guidelines for security considerations for all protocols.
Ten Oever added that other standards bodies including IEEE and ISO are taking ethical concerns and social responsibility into consideration when creating Internet protocols.
“This is complex,” ten Oever admitted. “We need to understand our own role and take responsibility. This does not mean that our technology is bad nor that it is good, but it definitely means that our technology is not neutral.”
He encouraged all IETF participants to review the draft guidelines for human rights considerations that is published as an information document on the group’s website. “We need protocol developers to road test the guidelines,” he said.
David Clark, head of the Computer Science Artificial Intelligence Lab (CSAIL) at Massachusetts Institute of Technology (MIT), shared with the audience that he has read the draft guidelines and found them both “fascinating” and in alignment with a movement he supports called “values in design”.
“Human rights are not absolute,” Clark said. “Designers of technology have a choice: to be in the conversation or not.”
As an example, he pointed to the Raven debate back in the year 2000, when the IETF declined to develop standards that would allow for lawful intercept of communications by law enforcement agencies.
“By declining, the IETF left the decision to others,” he said. “It doesn’t mean we didn’t get wiretap standards. It just meant they were made by someone else.”
Clark said the IETF has a choice: to continue designing protocols for a preferred outcome as it did in the Raven debate. Or it can incorporate into its protocol design a tolerance for a range of outcomes that it might not prefer, such as the ability to wiretap.
“You are designing the playing field, not the outcome of the game,” Clark pointed out. “But if you are clever enough, you can tilt the playing field.”
Following remarks from ten Oever and Clark, the IAB’s Lee Howard moderated questions from the audience.
Longtime IETF member Scott Bradner provided some background on the Raven debate for audience members who hadn’t participated in those discussions. “It was not an easy discussion,” Bradner said. “There were people who said it is a moral sin that governments wiretap, and also people who said it is a government’s responsibility to do so. It was a political discussion masquerading as technology.”
The audience seemed split on the IETF’s role in human rights discussions. Some members questioned whether the IETF was the right standards body to tackle these types of issues and warned that its engineering goals could be sidetracked by ethical debates. Others, including Clark and ten Oever, encouraged the IETF to take a more active role in these controversial debates than it has in the past.
“There are people such as lawyers, judges, legislators, public opinion, and the market that are thinking about ethics. This does not mean that we should not,” ten Oever said. “We cannot outsource our ethics to others and hope that they take care of it. But it also means that we are not going to replace them. We should, just within our little realm, take our responsibility.”
Similarly, former IETF Chair Harald Alvestrand argued that the IETF has no choice but to get involved in human rights-related debates and try to “tilt the playing field in the direction that we want.”
Daniel Gillmor of the American Civil Liberties Union likewise argued for the IETF to get involved in these debates because it is building important tools that everyone in the world uses.
“It is critically important that engineers, like all of us in this room, think ethically all the time about what the consequences are,” Gillmor said. “I’m really happy to hear that we are having this discussion and acknowledging that we are playing a role about whether people can exercise the rights that they expect to have on the Internet today, tomorrow, and the day after tomorrow.”
Howard ended the discussion by encouraging IETF participants to continue the conversation in the HRPC Research Group.
The technical development of “open standards” is the basis for many human rights. In this area the IETF and other standardization bodies could do far more. The openness of standards is more complex and misunderstood than the openness of standardization. Currently the IETF and most standardization bodies address openness of standardization. Designing adaptable standards would be a step forward. Please see http://www.isology.com/pdf/OpnStdsCallforAction.pdf for details.