extractfeed
A rolling, agent-readable changefeed for web scraping and data extraction.

Cloudflare shifts bot mitigation from risk scoring to continuous trust evaluation

Cloudflare announced a new approach to bot mitigation that moves from point-in-time risk assessment to continuous trust evaluation, introducing systems BotBase and Precursor to assess good and bad behaviors from bots and agents.

Markdown twin JSON

Anti-bot & Blocking Primary source release / significance 4

Briefing

Why it matters

This represents a fundamental architectural shift in how one of the largest anti-bot platforms thinks about traffic. Instead of classifying requests as good or bad at a single moment, Cloudflare is now modeling behavior over time, which could dramatically reduce false positives for legitimate automation while catching sophisticated adversarial bots that evade point-in-time checks. The introduction of Precursor Trace simulation also gives developers a way to test how their own automated agents are perceived, which may reshape how bot management vendors compete.

Sources

Watch next

Will other anti-bot vendors adopt continuous trust evaluation models, or will Cloudflare's approach create a new standard that competitors must match?

Topics: Cloudflare, BotBase, Precursor, bot-mitigation, trust-evaluation, behavioral-analysis, cloudflare, agentic-internet