Social Media Screening

Social media screening is the AI-powered analysis of an individual's public social media activity across Instagram, X (Twitter), TikTok, LinkedIn, and Facebook to identify risks such as hate speech, violence, substance use, and sexual content. Phyllo's screening replaces manual profile reviews — which typically take 2–5 days — with automated, multi-platform scans that deliver structured, GARM-compliant reports in approximately 15 minutes.

Phyllo's social screening requires name, email & education or employment details or social profile link to begin. The AI identifies an individual's profiles across five platforms, then scans all content types — text, images, audio, and video — simultaneously. Flagged content is classified across seven risk categories and compiled into a structured report with direct links to each flagged item, typically within 15 minutes.

Phyllo screens five major platforms: Instagram, X (Twitter), TikTok, LinkedIn, and Facebook. Each platform is scanned for text posts, images, audio, and video content in parallel — ensuring no content type is missed regardless of platform. Additional platform coverage is on Phyllo's product roadmap.

Phyllo's AI detects content across seven categories: hate speech, violence, sexual content, substance use, caption language toxicity, audio language toxicity, and visual attributes. The screening also includes adverse media checks across Google, news sources, & public govt websites for frauds. Organizations can define custom categories — such as competitor mentions, political commentary, or regulatory language violations — beyond these defaults.

Yes. Phyllo supports custom keywords as well as categories beyond the default safety categories. Organizations can configure specific terms, hashtags, topics, and themes relevant to their industry — for example, competitor brand mentions, role-specific risk criteria,adverse media checks or regulatory compliance language. The AI also detects keyword derivatives and synonyms automatically.

Manual social media review is slow, inconsistent, and legally risky — reviewers introduce personal bias, miss content across platforms, and produce no audit trail. Phyllo applies identical AI-defined criteria to every profile, covers five platforms simultaneously, analyzes all content formats including audio and video, reduces false positives to approximately 1%, and delivers auditable reports in 15 minutes instead of 2–5 days.

Each Phyllo screening report includes a platform-by-platform breakdown of flagged content, the detection category for each flag (hate speech, violence, etc.), the content type (text, image, video, audio), direct links to every flagged post, an adverse media summary from Google and news sources, and an overall risk classification. Reports are exportable in two formats, i.e. PDF and APIs suitable for compliance and audit records.

Phyllo's AI uses contextual, multimodal analysis — not simple keyword matching. The system evaluates intent by analyzing surrounding text, imagery, audio tone, and engagement patterns to distinguish genuine threats from benign content such as memes, satire, quotes, or news reporting. This contextual approach reduces false positives to approximately 1%, according to both Phyllo's internal testing & external validation.

Phyllo's standard screening returns results in approximately 15 minutes. The AI processes content across all five connected platforms in parallel with no manual review queue — compared to traditional manual screening workflows that typically require 2–5 days per individual and 40+ hours of analyst time.

Social media screening is legal for publicly accessible posts in most jurisdictions when conducted with individual consent, consistent application across candidates, and documented evaluation criteria. Phyllo supports compliant workflows aligned with FCRA (Fair Credit Reporting Act), EEOC (Equal Employment Opportunity Commission), GDPR (General Data Protection Regulation), and the EU-U.S. Data Privacy Framework. Organizations should consult legal counsel for jurisdiction-specific requirements.

Phyllo supports both one-time screening and continuous monitoring. The continuous monitoring feature provides real-time alerts when new flagged content is detected on a monitored individual's social profiles — useful for ongoing influencer partnerships, employee social media policy enforcement, or high-profile spokesperson management. The timeline of rerunning the report can be customised as low as to hours and as high as months/years. While rerunning it only evaluates new posts rather all screened earlier saving cost for the clients

Organizations can sign up for free to explore Phyllo's screening dashboard or APIs, download a sample BGV report to preview the output format, or schedule a call with Phyllo's team for a product walkthrough. Phyllo is available as a self-serve dashboard for individual teams and as an API integration for platforms embedding social screening into existing workflows — with typical API integration taking under 24 Hours.