Echoes & Algorithms is a mini-series delving into the evolving relationship between grassroots journalism and artificial intelligence (AI), exploring how AI influences narrative authenticity and sparking necessary discussions about the future direction of grassroots journalism organizations like Weave News amidst technological advancements.

Echoes & Algorithms: Attention, Autonomy, and the Future of the Feed

Artificial intelligence has quietly begun to rearrange the structures through which grassroots journalism is practiced, perceived, and absorbed. This is happening not only through its tools—automated transcription, summarization, image generation—but also through its more insidious role in shaping how audiences encounter the world: through feeds, recommendations, and optimizations. AI is not just an assistant to journalists; it is increasingly an invisible editor, a distributor, and at times, a gatekeeper.

At the heart of this shift lies a quiet, profound question: can we still choose what we pay attention to, or is that being chosen for us?

In an age of algorithmically curated content, the idea of cognitive sovereignty - the ability to direct one’s own attention, form judgments, and navigate information freely - feels increasingly fragile. And for grassroots journalism, whose power lies in amplifying overlooked truths and sustaining diverse public narratives, that fragility presents a serious threat.

The Disappearing Commons

News no longer arrives as a shared bulletin. It now trickles in through algorithmic tunnels: unique, personalized, increasingly opaque. While this has made some content feel more relevant, it has come at a cost: the erosion of a shared public sphere. We don’t all see the same story anymore. Sometimes, we don’t even see the same reality.

“If grassroots journalism exists to decentralize power, to tell the stories others won’t, and to serve communities rather than markets, then it must take seriously the systems through which it is seen and shared. It must contend with the algorithm not just as a tool, but as a terrain. We are not simply competing for attention—we are defending the capacity to direct it. Cognitive sovereignty may not yet be a widely used phrase, but it names something essential: the fragile freedom to think clearly, feel deeply, and decide what matters. Journalism, at its best, strengthens this freedom. In the age of AI, defending it may become our most urgent task.”

Policy institutions are taking notice. The German Ethics Council’s 2023 report emphasized that AI must enhance human potential, not replace it. Such declarations reflect growing awareness of how algorithmic systems, when left unchecked, threaten our capacity to reason independently. In journalism, that means grappling with the loss of a shared informational baseline, something grassroots journalism critically relies on to cultivate community-driven insight and accountability.

For grassroots journalism, which often exists to rupture consensus, to provoke critical thought, or to offer alternatives to dominant media narratives, this presents a fundamental problem. Stories meant to reach beyond the margins are buried beneath engagement metrics and behavioral predictions. The feed (now optimized for what will keep us watching) does not reward disruption. It punishes it.

The Ethics of the Feed

Tailored recommendation systems, while ostensibly neutral, are actually anything but. Built to predict what we’ll react to, they amplify what aligns with our biases, confirms our habits, or triggers our emotions. At a deeper level, they also help create the very things they supposedly reflect. In other words, they don’t offer discovery so much as reinforcement and a kind of social engineering that masquerades as objective reality. This is what makes them effective. And dangerous.

The result isn’t just more misinformation—it’s fragmentation. Even truthful reporting circulates in isolated pockets, shaped more by algorithmic design than public need. Informational environments become increasingly disjointed, curated less by free choice than by private incentives. In this climate, grassroots journalism faces an uphill battle. It must not only speak truth—but also break through the engineered silos of attention, across platforms optimized not for substance, but for stickiness.

Narrative Voice in a Predictive System

AI-generated content can assist with tasks like transcription, summarization, and even drafting. These tools promise efficiency and accessibility. But when deployed without care, they risk blurring what makes grassroots reporting distinct. Voices shaped by place, urgency, community, and resistance can begin to sound eerily the same—flattened by predictive language tools trained on dominant trends rather than divergent truths.

This is especially critical for emerging journalists. The process of developing a voice—one informed by lived experience, shaped by experiment, and honed through editorial friction—cannot be replicated by tools trained on consensus. And yet, these tools are often marketed as enablers of creativity. What they enable, more often, is conformity. To maintain narrative autonomy, it is not enough to use AI responsibly. Grassroots media must also ask what is lost when editorial tone is softened to fit algorithmic expectations. If surprise, challenge, or contradiction are penalized by platform logic, then those are precisely the traits we must defend.

Cognitive Sovereignty as a Public Good

People do not experience news in a vacuum. They encounter it in sequences—scrolling from a meme, to a protest clip, to a product ad. The context shapes the interpretation. Even truth, in this environment, is subject to framing. The more these sequences are tailored to individual behavior, the less stable public discourse becomes.

Cognitive sovereignty is not the same as intelligence or literacy. It is a condition—a set of prerequisites for thinking freely. First brought to public attention a few years ago with the resurgence of flip phones, the concept has increasingly been picked up by contemporary thinkers like Jack Lawrence in the realm of media ethics. It requires space, time, and some insulation from manipulation. But we are increasingly immersed in environments designed to erode it. Platforms harvest attention by bypassing our deliberation. We don’t choose the feed. The feed chooses us.

Grassroots journalism, then, must recognize its dual role: not only to produce truthful, independent content, but to help restore the very cognitive conditions that make truth legible. This includes advocating for structural reforms—not just ethical guidelines—that foreground transparency, resist algorithmic determinism, and prioritize audience agency.

What Can Be Done?

This is not a call for purity or technophobia. It is a call for awareness and agency. Grassroots organizations should consider the following:

Transparency around AI usage—not just disclosing tools used, but reflecting on how they shape output.
Resistance to algorithmic conformity—choosing voice and form that defy expectations when the story demands it.
Support for non-algorithmic channels—newsletters, community radio, local events, physical media.
Public education about how feeds work, how content is sequenced, and how attention is shaped.

Most of all, we must be vigilant against the slow normalization of predictive media environments. There is power in the pause, in the friction, in the refusal to be streamlined.

Reclaiming the Future

If grassroots journalism exists to decentralize power, to tell the stories others won’t, and to serve communities rather than markets, then it must take seriously the systems through which it is seen and shared. It must contend with the algorithm not just as a tool, but as a terrain. We are not simply competing for attention—we are defending the capacity to direct it. Cognitive sovereignty may not yet be a widely used phrase, but it names something essential: the fragile freedom to think clearly, feel deeply, and decide what matters. Journalism, at its best, strengthens this freedom. In the age of AI, defending it may become our most urgent task.