The anatomy of “fake news”


I found an online news article last week that made me wonder whether I was looking at fake news.  I discovered the article was originally written by Jessica Roy, a journalist at the LA Times, but the article had been altered to remove her name and change her words.  I confirmed with Jessica that her article had been plagiarized:

The article was almost identical. Almost. But if you read it closely, there were some quirks. The phrase “the alt-right” became “the alt-proper.” The magazine “Vanity Fair” became “Vanity Truthful.” The Free Speech Movement became the “Cost-free Speech Motion.” Instead of prompting people to follow me on Twitter, it asked people to “abide by me on Twitter.”

The rabbit hole goes much deeper.  Read all about it here: A fake news site stole my story. The Trump subreddit didn’t notice.

Fake news

What is it an how do you know it when you see it? Let’s ask the following questions about a possibly-fake news article:

  • was the article obviously generated by software?
  • is the content plagiarized?
  • are content sources and authorship redacted?
  • is the article hosted on a freshly-registered anonymous domain?
  • are the other articles on the website irrelevant and disorganized?
  • is there no apparent editorial board. physical office, or business?

If the answer to these questions is “yes” then you probably aren’t looking at real news.  Altogether, these criteria help to identify one type of fake news: bot news.  Last week I uncovered a bot newspaper site and worked with the LA Times to describe it.  This is the link to that article: A fake news site stole my story. The Trump subreddit didn’t notice.

CRTC provides new Net Neutrality framework

We had a recent discussion at IRN about net neutrality that involved the United States and the United Kingdom, which have approached the issue differently.  Now Canada, through the CRTC, have clarified their network neutrality rules.

From the Government of Canada:

The Canadian Radio-television and Telecommunications Commission (CRTC) today strengthened its commitment to net neutrality by declaring that Internet service providers should treat data traffic equally to foster consumer choice, innovation and the free exchange of ideas. As such, the CRTC today is publishing a new framework regarding differential pricing practices.

This framework supports a fair marketplace for services, cultural expression and ideas in which Internet service providers compete on price, quality of service, speeds, data allowance and better service offerings, rather than by treating the data usage of certain content differently.

The CRTC is of the view that differential pricing generally gives an unfair advantage or disadvantage to certain content providers and consumers.

The actual decision is contained within Telecom Regulatory Policy CRTC 2017-104:

In order to provide clarity to stakeholders, including consumers, content providers, and Internet service providers (ISPs), the Commission establishes a framework and sets out the evaluation criteria it will apply to determine whether an ISP’s specific differential pricing practice is or is not consistent with subsection 27(2) of the Telecommunications Act.

In this decision, the term “differential pricing practice” refers to the zero-rating or discounting of retail Internet data traffic. Practices associated with ISPs’ own managed Internet protocol networks are not included in the analysis and determinations set out in this decision.

The evaluation criteria are the following:

  • the degree to which the treatment of data is agnostic (i.e. data is treated equally regardless of its source or nature);
  • whether the offering is exclusive to certain customers or certain content providers;
  • the impact on Internet openness and innovation; and
  • whether there is financial compensation involved.

The decision appears to preserve a low barrier for Internet publishing by barring incumbent ISPs from creating pricing structures that could stifle online competition.

Interesting article: Intelligence report on DNC hack fails to establish clear attribution

Interesting piece on the flawed intelligence report on the DNC hack:

Next meeting: Nov 2, 4pm – Ian Miller on Networks, Complexity, and Intelligence

Join us for our monthly seminar meeting with a presentation on collective intelligence by Ian Miller followed by a group discussion.

Ian described his research as follows:

I take the perspective that social phenomena can be modelled with the use of computational agents, and that psychology provides models of agent behaviour.  Agents may be connected through a variety of networks that provide mediated vectors of transmission. In the particular case of online information networks, we observe nonlinear transmission phenomena in which memes (a unit of cultural reproduction) spread as a social information epidemic.  Online memes are a mechanism for agents to message and influence one another, providing a social coordination function. My work therefore uses social agents who transmit memes to model and explore social phenomena.


In this talk, I will discuss several recent studies, including:

– 2016 US presidential candidates and the memes created about them

– urban legend transmission

– the influence of Individuals and Organizations in open source software

– a computer simulation of the human population

– the research literature of social psychology and artificial intelligence




We are an interdisciplinary network of researchers with a shared interest in topics concerning the relationship between the Internet and society. Because more and more of our daily lives are affected by technology, technological development and social change are increasingly intertwined. Addressing questions around these changes in turn requires expertise from more than one discipline. That is why our aim is to share expertise, give each other feedback on research projects or ideas and establish dialogue that reaches across traditional disciplinary boundaries.

Our members include scholars from computer science, information science, psychology, sociology and political science and we currently hold monthly seminar meetings where people present finished projects or ask for feedback on initial ideas. These meetings are open to anyone interested and will be announced on the website.