Resources to help journalists and newsrooms confront power imbalances by promoting industry accountability and public understanding

Today, I am launching Algorithmic Literacy for Journalists (ALFJ), a collection of practical introductory resources to encourage and enable critical reporting on algorithms and their social impacts. 

The ALFJ website was developed with the support of the Reynolds Journalism Institute and in consultation with experts in journalism, computer science, and media literacy. The resource is available for journalists and newsrooms to use at no charge.

What’s an algorithm? Why bother with algorithmic literacy?

Algorithms—sets of instructions for performing a task or solving a problem—are the essential building blocks for a host of artificial intelligence (AI) systems revolutionizing journalism. Used to classify, prioritize, and filter information (including, for example, content recommendations by search engines and social media feeds), algorithms are reshaping how news is gathered and produced, distributed, and consumed.

Every news worker, regardless of specialty, needs some degree of algorithmic literacy, or the ability to understand and critically evaluate algorithmic systems, and the consequences of their use. 

Every news worker, regardless of specialty, needs some degree of algorithmic literacy, or the ability to understand and critically evaluate algorithmic systems, and the consequences of their use. 

Promoting algorithmic accountability

Algorithmic Literacy for Journalists includes six core modules. Three of these focus on algorithmic accountability reporting, journalism that informs the public about algorithms’ benefits and risks, and the AI systems they drive.

Algorithmic accountability reporting adapts journalism’s traditional watchdog function to investigate algorithms, their social impacts beyond the world of tech, and their potential for perpetuating existing biases, inequalities, and other social risks.

This section includes modules on relevant questions to ask, featuring sources from outside the industry, and the framing of news stories about algorithms and AI.

Responding to algorithmic “gatekeeping”

Three additional modules focus on “algorithmic gatekeeping” and practical responses to it.

Gatekeeping refers to media processes that filter—and potentially marginalize or block— information that reaches the public. It was first studied in the 1950s by communication scholars interested in how newspaper editors decided what stories to feature.

In current usage, algorithmic gatekeeping refers to a host of practices employed primarily by Big Tech companies rather than news editors. These practices impact which news stories circulate widely and which are buried from the public’s view.

ALFJ not only explains how shadow bans, advertising “blocklists,” and other forms of online content reduction can restrict journalists and newsrooms from reaching wider audiences, but it also provides resources for mitigating the impacts of algorithmic gatekeeping. In this section, the modules on shadow banning and advertising blocklists will be most useful to news managers, while the module on conducting DIY algorithm audits is intended for reporters interested in countering the journalistic pitfall of treating algorithms as inscrutable “black boxes.”

A gateway to additional resources, a clear source for key terms

The six modules focused on algorithmic accountability reporting and gatekeeping are ALFJ’s core elements, but two further features supplement them.

An annotated list of additional resources provides links to external sources—including more detailed how-to guides, a selection of articles that exemplify algorithmic accountability reporting at its best, and an entertaining, instructional online game— for those with the time and interest to dig deeper, beyond the essential elements provided by ALFJ itself.

An accompanying glossary provides clear definitions, without jargon, of key terms that are often used in discussions about algorithms and AI but not always explained.

Why is this vital now?

“When algorithms fail, they can lead to discrimination, financial losses, privacy breaches, and more,” the Nieman Journalism Lab reported in a 2019 article that described “the algorithm beat” as journalism’s “next frontier.”

Since then, those concerns have multiplied as AI systems driven by algorithms have become more widespread and powerful, with a corresponding increase in public concern.

During the same time, however, the “algorithm beat” has also developed, spurred by initiatives too numerous to name here, including the Pulitzer Center’s AI Spotlight Series, Northwestern University’s Generative AI in the Newsroom (GAIN) project, and MIT’s AI Risk Repository. The Solutions Journalism Network’s curated database of rigorous reporting on algorithms, artificial intelligence, and machine learning documents some of the fruits of these efforts.

Given all this, do we need even more reporting on algorithms and their social impacts?

“Yes, indeed,” is the unequivocal reply.

The general public continues to lack essential knowledge about algorithms, even as they rely on platforms such as TikTok, which use algorithms to make content recommendations, as important sources of news. One key takeaway is that journalists and newsrooms have vital roles to play in helping the public understand the promise, limitations, and risks of algorithmic technology in our everyday lives.

But promoting algorithmic literacy for journalists isn’t only about improving public understanding of algorithms or even recalibrating how news is digitally distributed and consumed, as important as those projects are. A

t a time when many Americans have difficulty distinguishing between news sources that do their own reporting and those that simply circulate already existing stories, ALFJ’s second key takeaway is that reporting of this kind can help revitalize public trust in journalism.

As powerful as they may be, AI systems are not designed to seek truth or ensure that their representations of the world minimize harm. By contrast, journalism’s ethical principles, including its cardinal commitments to transparency and accountability, distinguish journalists, newsrooms, and the profession.

As powerful as they may be, AI systems are not designed to seek truth or ensure that their representations of the world minimize harm. By contrast, journalism’s ethical principles, including its cardinal commitments to transparency and accountability, distinguish journalists, newsrooms, and the profession.

Algorithmic Literacy for Journalists was designed to encourage more news reporters, editors, and outlets to adopt that perspective, and especially to prepare them to cover current and emerging AI systems and the algorithms that power them accordingly.

Have feedback? Interested in getting more involved?

Please be in touch! I will continue to develop and update these resources, and I welcome your feedback on any aspect of the ALFJ website that you believe might need updating, revision, or expansion.

I am especially keen to connect with journalists and newsroom managers who might be interested in using one or more of the ALFJ resources to produce original reporting or to improve the reach of your work on social media. I would welcome the opportunity to discuss potential collaboration based on ALFJ resources. 

---

Cite this article

Roth, Andy Lee (2024, Feb. 24). Introducing Algorithmic Literacy for Journalists. Reynolds Journalism Institute. Retrieved from: https://rjionline.org/news/introducing-algorithmic-literacy-for-journalists/