External Reference — Archived 2026-03-27

The Future of Our Information Environment

Meeting people where they're at in an age of content overload — how young people actually consume political information and what it means for democratic communication

Katie Harbath · Published October 9, 2024 · Anchor Change (Substack)
Archived Reference. This is a summary and annotation of an external article by Katie Harbath, archived for Kyanos research. Original published at Anchor Change on Substack, October 9, 2024. Katie Harbath is a former Facebook policy director and a practitioner in political technology. This article draws on a Jigsaw and Gemic study of 52 young adults (18–24) in Bangalore and New York City.

Core Argument

Digital platforms and political communicators must adapt to how young people actually consume information rather than expecting users to engage with content on traditional terms. Resisting this reality doesn't protect information quality — it reduces influence over shaping the future information environment. The question is not whether to meet audiences where they are, but how to do so while maintaining the integrity of what is being communicated.

Central Finding

Young people (18–24) no longer rely on institutional credibility or academic credentials to assess trustworthiness. They prioritize perceived relatability and conduct their own research. "Information no longer needs to be based on institutional approval to be considered trustworthy." Trust is now personal, not institutional — and the communication infrastructure designed around institutional trust is becoming progressively less effective for this demographic.

Research Findings

Finding 1
Information Overload as Emotional Management Problem
Participants reported feeling emotionally exhausted by constant information pressure. Political and war content carried "social consequences" — a sense of obligation to engage — while entertainment content felt emotionally restorative. Users rapidly switch between content modes based on emotional equilibrium needs, not information-seeking goals. Most time is spent in "obligation-free" and light-content modes. The implication: political information competes with emotional restoration, not just entertainment.
Finding 2
Trust Decoupled from Institutional Authority
Young adults assessed trustworthiness through relatability and personal verification rather than source credentials. This shift undermines communication strategies built on credential-signaling (expert endorsements, institutional affiliation, academic validation). Content that feels authentic and personally verifiable outperforms content that asserts authority. For political communicators, this means the "who is saying this" signal is less effective than "does this match what I'm observing."
Finding 3
Politicians Rank Last on TikTok Follow Behavior
Per Pew Research data cited in the article, politicians' accounts rank at the bottom of account types followed on TikTok. Traditional information literacy models assuming linear decision-making — awareness, interest, evaluation, decision — do not match observed behavior in this demographic. The distribution channels that once reliably carried political information to young voters have fragmented or been abandoned by their intended audience.

Harbath's Prescription

Rather than dismissing young people's preferences, meet audiences where they are — using AI-powered content summarization and exploring new delivery mechanisms that respect user preferences. Podcasts are effective for Gen Z; traditional media reaches older generations. The underlying principle: format and channel adaptation is not a compromise of message integrity; it is the precondition for the message reaching anyone.

Kyanos Annotation: Applicability to Our Work

This article is the audience-side complement to Hall's supply-side analysis. Hall documents what AI systems retrieve about candidates; Harbath documents how the people those systems serve actually consume and evaluate political information. Together they frame the complete information environment challenge progressive candidates face.

The trust decoupling finding has direct REMEDY implications. If young voters assess trustworthiness through perceived relatability and personal verification rather than institutional authority, then third-party citations at Wikipedia and Ballotpedia — the backbone of REMEDY's source hierarchy — may perform differently across age cohorts. For voters who trust institutional sources (older demographics), Wikipedia/Ballotpedia citations are credibility signals. For voters who trust relatability, the candidate's own authentic voice may outperform institutional corroboration. This doesn't change the REMEDY source hierarchy (AI systems still weight institutional sources heavily) but it matters for how REMEDY outputs are explained to campaigns.

The emotional exhaustion finding explains low-engagement AI queries. Voters in emotional-management mode — which is most of the time — are not seeking comprehensive political information when they ask an AI chatbot about a candidate. They want a quick, low-friction answer that resolves uncertainty rather than deepening it. This is why APS measures whether AI responses are accurate and complete, not whether they are persuasive. A confused or wrong AI answer encountered during a low-engagement information-seeking moment is unlikely to prompt further research — it becomes the answer.

The channel fragmentation context validates AI as a priority channel. Politicians rank last on TikTok. Traditional media reaches a shrinking demographic. Organic search is declining as a political information source. Conversational AI is one of the few growing channels — and unlike social media, it is relatively early in the adoption curve. The "lapped" framing from Harbath's 2026 piece applies here: the window to establish a strong AI presence is open now because not all campaigns have started.

Archived for Kyanos research use  ·  Original article: Katie Harbath, Anchor Change (Substack), October 9, 2024  ·  All content is the author's own  ·  Annotation by Kyanos team