A New Map for Government Transparency
In a welcome step, Google has released important data that provides insight into the extent to which governments ask or require Google to provide private communications information to the government, or take down information that is made public on a Google service. To organize this data, Google launched the Government Requests Tool to display certain limited information about two kinds of government requests that Google receives, broken down by country:
- Content removal requests –search result removals and content removal from a variety of services (YouTube, Blogger, Groups, Orkut, etc.)
- Data disclosure requests – user identifying information, email logs, email content, etc.
The tool shows how many of each kind of request Google received during the last half of 2009 in countries where the government made at least a certain threshold number of requests in the specific category. With respect to content removal requests, the tool also tells us the percentage of requests Google complied with, and breaks down the number of requests further by product – for example, distinguishing between content removed from YouTube, Blogger, or from search results.
The data is not comprehensive, however, which Google explains in its FAQ: The statistics on content removal requests excludes requests related to child pornography or copyrighted content. The tool also does not cover direct government action to block services entirely in a country. It does, however, encompass requests related to a range of content like alleged defamation, hate speech, and impersonation, provided the requests come from governments and courts, as opposed to private parties. With respect to data disclosure requests, the data primarily covers requests from law enforcement on criminal matters, but excludes requests where Google is prohibited by law from even disclosing receipt of a request, as with National Security Letters under U.S. law. In addition, the statistics merely count the number of requests Google receives, even though a single request may ask for removal of multiple pieces of content or cover data from multiple users.
Given these limitations, it is difficult to draw certain conclusions from the data. For example, according to the tool, Google complies with around 80% of content removal requests from the U.S. and Brazil, and only around 53% of those from Spain and Australia. A number of factors could be at play in explaining the differences here, including differences in national free expression traditions, aggressiveness of law enforcement tactics, or divergences in whether the national law imposes liability on intermediaries like Google for third party content. Without greater granularity, it is difficult to assess certain causal relationships or extract trends.
While the data presented is neither perfect nor comprehensive, however, this tool is a promising first step towards increasing transparency around government practices that are often opaque, arbitrarily applied, and difficult to track from outside a company. Increasing transparency around these government actions can create important ripple effects that help keep the Internet open and free: Governments will be more reluctant to censor content or send tenuous or illegitimate user data requests when they know such requests are tracked and published. Users are better able to understand the privacy risks they might face online. Internet advocates can more effectively promote government practices that minimize the harm to privacy and expression. And citizens will gain valuable insight into actions their government may be taking that negatively impact their ability to speak online, access information, and protect their privacy – and act to hold their government accountable.
Increasingly, governments around the world are looking to technology companies as convenient points of control on online content or unlawful behavior, sometimes in pursuit of legitimate policy goals and sometimes aimed solely at stifling dissent or maintaining political control. In launching this tool, Google has provided a welcome example of how technology companies can help advance free expression and privacy on behalf of their users, and work to mitigate the human rights risks its products or services may present. We applaud Google’s initiative and hope other companies take similar steps towards increasing transparency around a range of government requests. If such transparency becomes the norm, users would have a better idea about the extent to which particular governments are requiring companies to take down public information or provide the government with private communications information.
In acknowledging the limitations of the Government Requests tool, Google expressed an intent to improve the tool over time, including, perhaps, providing greater granularity in its reporting on user data requests. We look forward to providing suggestions for how to maximize the impact and effect of this important resource.