Israel Buys Social Media Monitoring System That Can 'Plant Ideas' in Online Discourse
Software manual suggests that main use of the system isn’t to monitor individuals but to learn about trends and public sentiment; the real question is what the government will do with it.
The Israeli government has purchased a software system enabling it to monitor social media in general and specific users in particular. The bid, which was won by a company called Buzzilla, specifies that the software must have the ability to “plant an idea in the debate on social networks, web news sites and forums,” reports Ido Kenan on the website Room 404.
The main purpose of the software is to monitor debate on the internet and identify trends and feelings among the public. “From time to time, the ministries have the need for monitoring services, and recovery and processing of data on internet,” the bid request states. “These services are necessary for a range of needs in the government sector, such as generating useful information for the sake of ongoing activity, feasibility testing, identifying trends, identifying needs, and identifying and handling crises.”
The Finance Ministry, which issued the request, further explains that until now, ministries requiring of such services had obtained them from different sources, so it decided to find a system that can supply all of the ministries’ needs.
The problem, writes Kenan, lies in the optional abilities of the system, which include enabling the government to plant ideas in conversations on social networks and forums through an automated or semi-automated mechanism. As Kenan points out, the authors of the proposal request could have written “introduce an idea,” suggest it, present it or some other synonym. Instead, they wrote of the requirement to “plant” the idea using an automated or semi-automated mechanism – meaning without any, or with little, further human intervention. That does not smack of open, transparent discourse, he writes.
In Kenan’s view, the system is a propaganda machine, but former technology correspondent Nitzan Weidenfeld shrugs that the system is a standard one offered by social analytics companies like Brandwatch – “the idea is that you can post something from the system directly.”
“The way they wrote is unfortunate,” Weidenfeld adds.
Indeed, the Buzzilla user guide makes that ability sound more like Weidenfeld puts it and less like Big Brother: a tool designed to enable reactions from the system itself. “Organizations that monitor the discourse in order to locate opportunities to take part in discussions online can use this module to manage the participation… the module also enables monitoring the volume of reactions, and a breakdown of the sites where intervention is carried out,” the guide states.
The content monitoring tool can also generate reports on trends, analyze sentiment on specific topics, carry out complex searches and cross-reference information. Users can type in parameters in free language or build complex queries to achieve more precise results.
The Buzzilla manual indicates that the main use of the system isn’t to monitor specific individuals but to learn about trends and public sentiment, which is not necessarily a bad thing – for instance, the government could gain insight into how the public feels about certain policy moves. The real question is what the government is doing with the system.
There is the possibility of swaying an existing debate, which is worrying; but as the system also offers a breakdown of user, even more worrying is the theoretical ability of the ministries to focus on specific ones, such as supporters of the boycott, divestment and sanctions movement against Israel.
These abilities are not at the center of the government’s request, nor are they stressed in Buzzilla’s own literature. The Finance Ministry announcement provides no ethical restraints on collecting or using data. However, it may bear adding that the entry screen to the system does not use encryption (https) but an ordinary connection, which could help hostile elements.
When asked whether there are constraints on using the system to monitor specific people or if government employees are required to identify themselves when they participate in online discourse, the Finance Ministry replied that the individual ministries buying the service are responsible for setting its own policies.
Attorney Yehonatan Klinger, the legal adviser of the Digital Rights Movement, points out that under the law protecting individuals' right to privacy, any collector of information has to advise the subject of the data whether he has the duty under law to deliver the information, and for what purposes.
When a public body is involved, information may be transferred between government bodies only when it is necessary for their legal operations. "In our case, since there is no legal obligation to deliver information, and since there is no certainty regarding the ways this information will be transferred and processed, there is another problem: What's going to be done with that information," he says.
In practice, the fact that a person publishes anything online does not enable somebody else to collect that information and use it, he says. When individuals posts on social networks, they have a reasonable expectation that their friends will use it, but not that it will serve potential employers, for example.
The state has twice argued that the data is already available on internet, so legislation isn't necessary. The law forbids a person to make use of information for purposes other than what it was intended for, says Klinger. A legal precedent was made in the case of an insurance company: The court ruled that it couldn't use information received for Purpose A to fulfill Purpose B. The U.S. courts did the same in the case of U.S. vs. Jones, which reached the Supreme Court in 2012 – protracted collection of information on a person using technology requires judicial permission.