SODP Dispatch - 11 December 2025

Down-ranking Polarizing Content Lowers Emotional Temperature on Social Media

By Tiziano Piccardi

Reducing the visibility of polarizing content in social media feeds can measurably lower partisan animosity. Researchers developed an open-source web tool that reranked feeds of consenting participants on X in real time for 10 days before the 2024 U.S. presidential election.

Using a large language model, researchers identified posts likely to polarize people—those advocating political violence or calling for imprisonment of opposing party members. These posts weren't removed; they were simply ranked lower, requiring users to scroll further to see them. The results: reducing exposure measurably improved participants' feelings toward people from the opposing party and reduced their negative emotions. Importantly, these effects were similar across political affiliations.

Feed algorithms are typically optimized to capture attention, and as a result, they significantly impact attitudes, moods, and perceptions of others. Progress in large language models now enables platforms to detect polarizing content and give users more control over what principles guide their feeds.

Read More ▸

🎁 The 2025 Gift Guide for Journalists | Nieman Lab

Nieman Lab's annual gift guide features everything from Field Notes' Front Page reporter's notebooks ($16.95 for two) to the Oxo Brew Rapid Brewer ($44.99) that makes cold brew concentrate in five minutes instead of 24-hour steeping. For tech organization, the Capacities app offers daily notes and custom object types for tracking interviews (free tier available). News Tower ($24.99 on Steam), an indie video game set in 1930s New York, lets you manage a newsroom while navigating mafia bosses and Tammany Hall cronies. Also featured: Jo Piazza's thriller Everyone Is Lying To You ($19) with a Ballerina Farm-inspired influencer and journalist protagonist, Kurt Vonnegut's surprisingly timely 1952 novel Player Piano ($19) about an automated society, and Stacy-Marie Ishmael's free newsletter The Main Event that pairs personal reflections with poetry recommendations.

Read more ▸

By Vahe Arabian

📈 AdTech Trends for 2026: Privacy-First Strategies & AI-Powered Monetisation

As we look toward 2026, the AdTech landscape is shifting, driven by privacy-first approaches, AI-powered automation, and intelligent frameworks focused on building trust-based monetisation strategies.

Here's what publishers and advertisers should keep an eye on to stay ahead:

1. Privacy-First Monetisation: Consent as a Revenue Driver

Your consent strategy isn't just a compliance issue; it’s directly tied to your monetisation strategy. The way publishers track consent at the granular request level is becoming crucial, as it influences yield outcomes and programmatic bids. Granular consent tracking and ensuring proper data authorisation will unlock higher CPMs and better monetisation results.

Actionable Tip: Bring consent management into revenue operations and ensure your consent framework is optimised for monetisation.

2. Multi-ID Strategy Is Essential

As we move past the decline of third-party cookies, a multi-ID strategy will be critical. No single identifier has universal adoption across DSPs and demand partners. Publishers will need to select identity frameworks that suit their needs, whether warehouse-based, ID spine, or CDP-based solutions.

Actionable Tip: Evaluate the right identity solution based on your publisher type (e.g., large enterprise, subscription-first) and focus on addressability to ensure efficient monetisation.

3. AI Automation Takes Over Optimisation and Creative

AI isn’t just about automating campaign optimisation; it's moving into creative production, powering everything from ad creatives to content optimisation. By 2026, expect AI-powered tools to handle end-to-end programmatic operations, freeing up human resources for higher-level strategy.

Actionable Tip: Embrace AI tools to automate creative development and campaign optimisations, while focusing on strategic decision-making.

4. Contextual Targeting Becomes a Privacy-First Alternative

As privacy regulations tighten, contextual targeting will see a renaissance. AI-powered contextual engines will enable precise targeting based on content rather than user data, allowing advertisers to deliver relevant ads without compromising consumer privacy.

Actionable Tip: Invest in contextual targeting technologies that leverage AI to deliver privacy-respecting ads based on content relevance and audience sentiment.

5.  Connected TV (CTV) Maturity and Programmatic Expansion

By 2026, Connected TV (CTV) will reach new heights in programmatic sophistication. CTV advertising will evolve beyond simple audience targeting to embrace data-driven creative optimisation, cross-platform orchestration, and real-time bidding. As streaming platforms mature, CTV advertising will seamlessly integrate into broader omnichannel campaigns, ensuring advertisers can target audiences across web, mobile, and TV in a unified experience.

Actionable Tip: Integrate CTV into omnichannel strategies for a unified ad experience, and adopt programmatic tools that support real-time optimization and cross-platform targeting.

6. Trust as a Core Asset

Building trust with both consumers and downstream vendors is essential. As more data privacy regulations come into play, transparency in how data is collected and used will become a competitive advantage for publishers.

Actionable Tip: Focus on building trust through clear data practices, proper consent management, and transparency with advertisers and consumers alike.

Looking Ahead to 2026

The future of AdTech is privacy-first and AI-powered, with an increasing emphasis on trust-based monetisation strategies. Publishers who succeed will be those who:

• Optimise their consent strategies to drive revenue.

• Adopt multi-ID solutions for better addressability.

• Leverage AI for creative and campaign automation.

• Adopt contextual targeting to respect privacy while delivering precision ads.

• Integrate CTV into their omnichannel advertising strategies.

The key to thriving in 2026 is adaptability and strategic foresight. Those who embrace privacy-respecting, AI-driven frameworks will lead the charge.

European Commission Opens Google AI Investigation | Press Gazette

The European Commission launched an antitrust investigation into Google over use of publisher content for AI products "without appropriate compensation." The probe examines whether Google breached EU competition rules by imposing unfair terms or granting itself privileged access to publisher and YouTube content, potentially abusing its dominant position.The investigation follows complaints from Movement for an Open Web, Foxglove, and the UK's Independent Publishers' Alliance. Publishers cannot block Google's AI bots without being removed from search results altogether. James Rosewell called it "double daylight robbery, stealing content from publishers to inform their models and then using these outputs to steal traffic from them." YouTube content creators must grant Google permission to use their data for training AI models without remuneration, while rival developers are barred from using that content.

Read more ▸

First-Party Data Becomes Publisher Lifeblood | INMA

Publishers are rewriting playbooks as third-party data crumbles. For the Financial Times, first-party data is one of only two company-wide OKRs. Der Spiegel required app registration and saw an initial user dip but significant boost in quality, enabling direct subscription sales without app store intermediaries. Condé Nast evolved from mass registration to brand-level strategies. Newsletters remain the biggest sign-up driver, with sophisticated tactics like evolving "save" functions into wishlist tools. The business case: the gap between low-value, untargeted ads and high-value targeted ones can be 10x. Among all data points, age is emerging as the most potent predictor. Key takeaway: Quality beats quantity. A smaller pool of identified users is often more valuable than a large anonymous audience.

Read more ▸

Comment Sections: The Untapped Gold Mine | AdExchanger

Kelly Andresen, former president of national sales at USA Today, joined OpenWeb as EVP of demand sales. The platform builds, moderates, and monetizes website comment sections. Comment sections remain an untapped gold mine of intent-based data, similar to Reddit's conversations between real people. OpenWeb places ad inventory inside comment threads and shares revenue with publishers. The platform uses five layers of moderation, including AI handling 90%, with humans on the final layer. Community Exchange launched as an ad targeting platform powered by anonymized data from 100 million+ logged-in users. The value: proving a person commented about a product or topic shows intent closely aligned with purchase behavior—something difficult for content companies to prove.

Read more ▸

Photographer Documents Disappearing Newsrooms | Poynter

Ann Hermes is documenting local newspaper newsrooms while they're still around. She's visited 50+ newspapers since 2016, focusing on family-owned papers. Her work was highlighted in The New Yorker last week. In newsrooms, she sees familiar echoes: police scanners, maps, newspaper morgues no one has time to digitize. One surprise: local newsrooms aren't microcosms of national ones. "There's a much more nuanced dynamic going on that I had not fully appreciated. They very much reflect their communities." Hermes said the decline is well-documented, but seeing it is something else. "The sliver of time is closing on this. It's a window of opportunity that is closing." The spaces are changing due to remote work, shrinking staffs, and costly downtown spaces.

Read more ▸

Mentions, Citations, Clicks: Your 2026 Content Strategy | Search Engine Land

Generative systems like ChatGPT, Gemini, and Perplexity are taking over early discovery. Visibility now isn't just about who ranks—it's about who gets referenced inside the models. Siege Media's study of 7.2 million sessions shows pricing content saw strongest growth while top-of-funnel guides declined sharply. But every major content category saw increased engagement as users reach sites later in the journey, already motivated. Seer Interactive found organic clicks rose from 0.6% to 1.08% when sites were cited in AI Overviews. ChatGPT had 16% conversion rate versus Google organic's 1.8%. The 2026 strategy requires two approaches: For the agent (execution), ensure BOFU content has clean schema and accessible APIs. For the human (selection), double down on TOFU with focus on mentions and citations.

Read more ▸