March 2026, Arbiter Social Intelligence Reports + Sakhi at Psychology of Tech

This month, we are introducing Arbiter Reports — web-docs that provide structured, high-level insight into how online conversations form, evolve, and shape public discourse across issues. These reports offer a clear, at-a-glance analysis of emerging narratives, helping you quickly understand how discourse is developing before diving deeper into detailed case studies generated by Arbiter's Agent on the platform.

What began as a mass-casualty event at Bondi Beach did not remain a local story online. On Twitter, it was quickly framed within a broader geopolitical and identity-linked conflict. On YouTube, it stayed centered on crisis response and community resilience. Arbiter's analysis shows how platforms shaped the meaning of the event in different ways. Twitter as the Political Arena (125.2M interactions in top theme): The dominant narratives were "Violence Against Jewish Communities" and "Geopolitical Tensions and Responses," positioning the attack within a wider political and international context. YouTube as the Response Hub (84.9M interactions in top theme): The leading theme was "Crisis Response and Community Resilience," emphasizing institutional updates, public safety, and societal impact. The focus moved quickly from reporting the attack to debating immigration, national security, and government accountability. On Twitter, the story expanded into geopolitics and identity-linked conflict. On YouTube, it remained anchored in response and resilience. The same event produced different public meanings.

The China–Philippines tensions in the West Philippine Sea did not unfold as a steady diplomatic debate online. Arbiter's analysis shows attention clustered around a few high-impact bursts — with YouTube carrying the largest share of engagement. YouTube as the Engagement Engine: Top themes included "Cultural Narratives and Dynamics" and "Viral Content and Memes." High-reach content leaned toward creator-style explainers and broadly shareable framing rather than strict maritime policy discussion. Twitter as the Diplomatic Stream: Themes centered on "Geopolitical Tensions and Responses," "Chinese Aggression in EEZ," and maritime governance — framing the dispute in sovereignty and security terms. Late January spikes were driven by general China-related viral content, while early February surges aligned with coast guard confrontation narratives. Across both platforms, a small number of breakout posts accounted for most interactions. The dispute is constant, but digital attention is episodic. When viral framing aligns with confrontation, the narrative accelerates — and YouTube drives reach.

A single image — Amazon's electric vans charging via diesel generators — generated 859,224 interactions and became the symbol of a growing credibility crisis in climate policy. Our multi-platform analysis reveals how skepticism is outpacing optimism in the AI and sustainability debate. Data center sustainability: Narratives are mixed but lean positive. Tech investors and environmental advocates emphasize energy-efficient AI and sustainability initiatives, while critics raise concerns about water usage and broader ecological impacts. Policy support for renewable energy remains strong, with narratives around "Green Job Growth" and "Local Solar Projects" gaining traction. The dominant counter-narrative is about Climate Skepticism. Posts questioning "Net Zero Feasibility" and exposing "Greenwashing" are dominating the discourse, with minimal pushback from environmental advocates, indicating "Strategic Silence." The clean energy movement is losing the narrative battle on social media. Without clearer communication and evidence from policymakers, skepticism will continue to grow.

We also presented Sakhi at the Psychology of Technology Conference in Washington, DC last November. The full deck is now public on our research page. The lessons-learned slide is the one we keep coming back to.

First: design for collaboration, not replacement. Build workflows where AI assists, while human insight ensures relevance and accountability. The frontline health workers we build with are the experts in their communities, full stop.

Second: do not overhype AI. The participants will buy it and you will not be able to dial it back. We saw this firsthand when an early version of the prototype made claims about its capability that we then had to walk back in the field.

Third: conversational agents lack conversationality in the medical domain. Off-the-shelf LLMs answer the question but miss the conversation around it. We built explicit human-in-the-loop hand-offs to frontline health workers as a result.

Fourth: lack of contextual awareness really hurts the model. We are now investing in long-term memory and case-history retrieval as a result.