E-mail attackers never sleep

Achieving the top spot is often a reason to celebrate. For example, The Globe and Mail takes pride in having the most National Newspaper Award wins – 216 and counting. Sometimes, though, coming in first is a dubious achievement.

During the lead-up to the holidays, The Globe’s chief information security officer, Darrin Nowakowski, stopped by my desk. His team had been analyzing the frequency of e-mail attacks across the organization, using an artificial intelligence tool. During a 90-day sample period, I had received more attacks – the umbrella term for phishing, fraud, malware, scam, extortion and social engineering e-mails – than any other editorial staff member. (Although Ottawa bureau chief Robert Fife was a close second.)

A significant amount of human effort went into “training” the tool to recognize attack e-mails. For example, over the course of several months, staff were asked to forward suspected phishing e-mails for examination. Being a natural rule follower, I did this dutifully, not realizing that Darrin and his team were reading each one and “feeding” them to the AI. The goal is for the now-informed interface to identify and remove attack e-mails before staff open them and risk triggering a cyberattack.

The next step will be to train the tool to recognize abusive messaging, which is a significant source of distress for journalists. Globe staff are encouraged to report abuse to their managers and human resources, so protective measures can be taken – for example, blocking senders, sending a cease-and-desist reply from a non-personal company e-mail address, perhaps getting the police involved. If the sender is a subscriber, their subscription may be cancelled.

I’ve seen some of these e-mails, and they are disturbing, to say the least, berating journalists with ugly sexual slurs, calling them liars and accusing them of bias. One of the milder messages advised a reporter to “get lost.” Other examples are unpublishable because they contravene The Globe’s newsroom guidance regarding the use of obscene language. (In case you’re wondering, in the rare instances when an obscenity is permitted, it is because the circumstance of its invocation is newsworthy in itself.)

A 2021 Ipsos poll found that 72 per cent of the journalists and media workers surveyed had experienced harassment in the past 12 months, with 34 per cent reporting they had received threats or harassment online, including via e-mail, on a monthly basis or even more frequently. And among those who experienced online abuse, 84 per cent said the frequency had ramped up over the past two years.

Women journalists are targeted more frequently than their male colleagues, with 73 per cent saying they had experienced online violence, according to a UNESCO report. Eighteen per cent had been threatened with death or sexual violence.

Covering COVID-19 became an unexpected risk factor for journalists, with the pandemic “turning media workers into targets for the public’s frustration,” observed the Taking Care report on media workers’ mental health, well-being and trauma.

A second chapter of issue-related abuse of journalists is playing out now, as members of the public direct their shock, anger and fear over the terrible events taking place in the Middle East toward those covering the war. Globe journalists have been accused of both pro-Israel and pro-Palestinian bias – sometimes in the same article. Comment moderators have been accused of the same. This is known as the hostile media effect, which The Oxford Handbook of Political Communication describes as a phenomenon “where opposing partisans perceive identical news coverage of a controversial issue as biased against their own side.”

Authors of the handbook draw a connection between the hostile media effect and declining trust in news organizations – which, they say, “has important consequences for democratic society.” How so? Turning away from balanced coverage in independent media they perceive as biased, the authors say, people may seek out reporting that confirms their own beliefs, leaving them less informed.

Meanwhile, what happens to journalists on the receiving end of abusive e-mails and other forms of online violence? These messages of hate take a toll on journalists’ mental health. “The majority of people who experienced threats and harassment – whether online, in the field or in the newsroom – reported being psychologically harmed by them,” the Taking Care report noted.

Journalists are trained to face danger on the job, but they are also human.

During the 90-day period my colleague Darrin and his team analyzed e-mails with the help of the AI tool, Globe e-mail addresses received more than 7,000 attacks; a peak day (Dec. 7) saw 421. And while this analysis doesn’t differentiate between messages that came from bots or humans, we do have a protocol for responding to people who cross the line. We want feedback, including strong opinions, but please remember that a human being is on the receiving end of your e-mail before you click “send.”