This is the fourth in a series of extracts from a draft book chapter on ethics in data journalism. The first looked at how ethics of accuracy play out in data journalism projects, the second at culture clashes, privacy, user data and collaboration, and the third at mass data gathering. This is a work in progress, so if you have examples of ethical dilemmas, best practice, or guidance, I’d be happy to include it with an acknowledgement.

Protection of sources

Most news organisations’ professional guidelines include sections on protecting sources. In some countries this is also enshrined in law. Many journalists, however, are not aware of how they can betray sources’ identity by publishing original files online.

Metadata stored in those files – information about the date and location of access, the computers and accounts used, and other data, can be used to identify a leaker. Even photocopied or printed materials can bear invisible digital watermarks which describe what machines were used to produce them, and when (Reimer, 2005; PicMarkr, 2008).

Journalists should, therefore, avoid publishing original source material online where the identity of the source is not known. For the same reason, journalists should avoid providing such material to public authorities, and exercise caution around storing such files online in ‘cloud’ based services which may be subject to interception by security services or legal pressure.

In some cases lack of cooperation with public authorities may also represent a form of ‘virtue ethics’ in asserting the independence of the journalist or news organisation. The International Consortium of Investigative Journalists (ICIJ), for example, resisted sharing the results of their investigation into offshore banking with government agencies (Ryle, 2013; Guevara, 2013) because:

“The ICIJ is not an arm of law enforcement and is not an agent of the government. We are an independent reporting organization, served by and serving our members, the global investigative journalism community and the public.”

This is particularly interesting given the international nature of the organisation, where local claims on duty (de-ontological ethics) might be weaker than in a national or local news organisation (see for example White, 2008 – PDF).

Leaks and ‘war’

“Most people would accept restrictions on war reporting to prevent information reaching the enemy which would endanger military operations,” writes Karen Sandler in Ethics & Journalism (2003). The ethical principle in play here is of minimising harm, but, as Sandler points out: “Avoiding harm to UK nationals may involve causing harm to Afghan citizens.”

Dealing with ‘mega leaks’ that relate to the military or national security raise similar issues. When war is constant and the enemy ill-defined, as is the case in the ‘war on terror’, balancing ‘public interest’ against possible harm is particularly difficult. Intelligence strategist George Friedman notes of Edward Snowden, for example, that:

“[He] is charged with aiding an enemy that has never been legally designated. Anyone who might contemplate terrorism is therefore an enemy.” (Friedman, 2013)

Publishers might ask whether ‘warlogs’ containing historical about deaths in Iraq and Afghanistan weakened the security of US, UK or German military operations. But there is an equal consideration to be given to the ethics of not publishing, as Sandler adds:

“The difficulty arises when failure to report something would cause unjust harm to someone else and at the same time its publication would prejudice others. [In] cases where harm to the national interest cannot be proven, there must be a strong presumption in favour of openness and transparency to safeguard honest and effective government.”

This, of course, is the discussion taking place around Edward Snowden’s leaks about the surveillance of internet traffic by US, UK and other government agencies. Australian academic Ben O’Neill’s assessment of Snowden’s actions quotes legal scholar George Strong on the enforceability of contracts that:

“… an illegal contract is one that is unenforceable as a matter of policy because enforcement would be injurious to the best interest of the public.”

On the matter of database property rights he adds in a separate post:

“When a private firm commits a crime using its own property as an instrument of wrongdoing it loses the right to claim ownership as a safeguard against investigation … The same applies to documentary evidence of a crime — it may legitimately be taken by an investigator as a means of proving criminal wrongdoing …

“Government claims to ownership have no special status in this regard, and do not override these ordinary principles of property rights. In fact, the situation for government claims of ownership is even weaker than for a private enterprise, since the latter will generally have acquired the tools of its criminal dealings with its own money.” (O’Neill, 2013b)

The same might apply to publication and the disregard of any government notices not to publish.

In the next part I look at automation and feeds. If you have examples of ethical dilemmas, best practice, or guidance, I’d be happy to include it with an acknowledgement.