AGENTIC COMMONSAI industry briefings

繁中EN

TopicAI Chips & ComputePublished

All English articles

Cloudflare Impact donations and nonprofit support

On this page6 sections
  1. The milestone, in numbers
  2. Why the free tier was never charity
  3. The demand side is worse than average
  4. From defense to AI-era enablement
  5. The takeaway
  6. Sources

Summary: Cloudflare announced on September 30, 2026 that its Impact programs will reach $100 million in donated services this week. Project Galileo now covers more than 3,500 domains in over 120 countries, and expansion includes AI crawl controls, nonprofit credits, and automation for human-rights organizations. The author notes the figures and framing come from Cloudflare’s own announcement rather than an independent audit.

The milestone, in numbers

On September 30, 2026, Cloudflare announced that its Impact programs will reach $100 million in donated services this week. The programs protect journalism outlets, civil society groups, state and local governments, election bodies, and public schools from cyberattacks.

The scale is worth pausing on. Project Galileo, launched in 2014, now covers more than 3,500 domains in over 120 countries. In 2025 alone, Cloudflare reports blocking 38.5 billion attacks against those participants — about 105.4 million per day. The Athenian Project protects 440+ election websites across 33 U.S. states, expanded to eight countries. Cloudflare for Campaigns covers 530+ campaign sites.

Why the free tier was never charity

The part I find most instructive for builders is the origin story. Cloudflare started as a free service whose attack data fed paid products. They then noticed free customers doing essential work — reporting on corruption, covering the invasion of Crimea — were absorbing some of the largest attacks on the network.

That observation turned a growth tactic into a business pillar. Their own framing is blunt: the Impact programs were never philanthropy; they are part of the business and mission. For anyone designing a free tier, that’s the pattern worth studying — the free users who stress your infrastructure hardest may also be the ones whose protection defines your reputation.

The demand side is worse than average

Cloudflare’s first annual report on attacks against civil society, released in 2026, found these organizations are targeted more frequently and more intensely than typical customers. Project Galileo participants faced website vulnerability exploitation attempts at more than seven times the rate of an average user. Applications to the program are on pace to more than double year over year.

If you build security tooling, that’s a market signal about an underserved segment — and a reminder that attack volume is not evenly distributed across your customer base.

From defense to AI-era enablement

The programs are expanding beyond firewalls. Free Bot Management and AI Crawl Control now go to over 750 journalists and news organizations, giving them control over how AI crawlers access their content. The startup program — $250,000 in credits — opened to non-profits for the first time, with the first 30 organizations announced this week. Cloudflare engineers also built automation tools for three human rights organizations.

A concrete example: at a September 9, 2026 hackathon in Barcelona co-hosted with Media Party, the winning team AIdas built a tool comparing how different LLMs answer politically contested questions, publishing responses as open data. The themes — workflow automation, agentic journalism, synthetic-content verification, information integrity — read like a roadmap for what newsrooms actually need next.

This shift from “protect them” to “equip them” echoes what I noted when Cloudflare gave every agent branch its own runtime: the developer platform keeps becoming the delivery mechanism for the company’s public-interest work, not just its commercial one.

The takeaway

A hundred million dollars in donated services started as a data-collection strategy for a free product. If you’re deciding what your free tier is for, Cloudflare’s example suggests asking a sharper question than “what will drive conversions”: which users, if attacked or squeezed out, would define what your product stands for — and what would it cost to protect them?

One caveat: these are Cloudflare’s own figures and framing from its anniversary announcement, not an independent audit of impact.

Our take: Eligible public-interest organizations should explore the support available, while small product teams should wait before copying the whole program. Cloudflare says Project Galileo covers more than 3,500 domains in over 120 countries, but that reach does not establish that another company can sustain the same commitment. Assuming a team has limited engineering time, it should first determine what service and support it can afford to maintain. We would favor expanding a similar program if evidence showed that recipients received lasting protection and the team could cover its costs. Builders without a public-interest service goal can safely ignore this announcement.

Sources

AGENTIC COMMONSOperated by PHLEGON LABS
ShareXEmail
Support us

Related reading

  1. Cloudflare CryptoLabe for post-quantum discovery

    Migration teams should borrow CryptoLabe's evidence process; Cloudflare targets 2029 but lacks a ground-truth dataset.

    AI

  2. Cloudflare 2026 midterms: attacks came weeks early

    Cloudflare's midterm data: vulnerability probes outnumbered DDoS 180 to 1 on election sites, peaking weeks before the vote — harden year-round.

    Cloudflare

  3. Cloudflare Sees AI Pentest Traffic Rise 30x Since October

    Cloudflare reports AI pentest traffic up 30x since October 1, but the breaches it details ran on unscoped CI/CD tokens.

    Cloudflare