Cloudflare Launches Adaptive Intelligence to Combat AI-Era Bot Attacks in Real Time

Cloudflare launches 'Adaptive Intelligence' to fight AI-era bot attacks in real time

Cloudflare introduced Adaptive Intelligence for real-time cyberattack detection. This engine learns from global internet traffic to block emerging bot threats. It continuously retrains on live data, unlike slower traditional security systems. The technology analyses behaviour patterns to identify slow and evolving attack campaigns. Cloudflare has announced Adaptive Intelligence, a new continuous detection engine built into its Bot Management platform that is designed to detect and respond to automated cyberattacks in real time. The company said the technology uses insights from trillions of requests across Cloudflare's global network to continuously learn from live internet traffic and automatically generate short-lived rules aimed at blocking emerging bot attacks. The move comes as automated cyberattacks become easier and cheaper to launch. Cloudflare said modern AI tools, compromised-device networks and software that can mimic human behaviour have lowered the technical and financial barriers for attackers. Also Read: AI cyber risk is biggest immediate threat to global financial stability: FSB chair Andrew Bailey For online businesses, this creates an uneven playing field: attackers can make repeated, low-cost attempts, while security teams need to identify and stop threats immediately without blocking legitimate users. Cloudflare says traditional security tools are too slow According to Cloudflare, many conventional security systems continue to depend on fixed rules and scheduled updates, which can take weeks or months to deploy. Attackers, meanwhile, can modify their techniques within hours or even minutes. Adaptive Intelligence is designed to continuously retrain on live traffic rather than wait for periodic updates. "Building taller walls fails when the cost of scaling an attack is effectively zero," Dane Knecht, CTO at Cloudflare, said. "To stop modern bot threats, you have to flip the economics on the attackers." The company said its system creates a continuously changing security environment, making it harder for attackers to study defences and develop techniques that can be reused at scale. Also Read: OpenAI-Hugging Face incident exposes cybersecurity's 'human-speed' problem How Cloudflare's Adaptive Intelligence works Cloudflare said Adaptive Intelligence can analyse more than a trillion web visits every day to identify new attack patterns. The system can continuously integrate new bot frameworks and bypass techniques into its detection engine. It also generates short-lived, targeted rules that are designed to block specific threats and then rotate, forcing attackers to repeatedly adapt. The technology is also designed to detect attacks that develop slowly over longer periods. Cloudflare said it can analyse behavioural patterns across multiple timeframes to identify threats such as slow credential-stuffing and scraping campaigns, which can evade traditional detection systems. Another component of the system combines browser-level session behaviour signals from Cloudflare Precursor with telemetry from Cloudflare's global network. The company said this allows the platform to assess whether traffic is automated or abusive without relying only on basic tests that could mistakenly block legitimate users. Cloudflare said Adaptive Intelligence can also test security updates against live traffic before deploying them. The system verifies their accuracy and looks for potential false positives before updates are rolled out. The company said this allows security upgrades to be deployed autonomously without scheduled downtime.

Cloudflare Announces Adaptive Intelligence Detection Engine Built into Cloudflare Bot Management

Cloudflare, Inc. announced Adaptive Intelligence, a new continuous detection engine built directly into Cloudflare Bot Management. Powered by insights from trillions of requests across Cloudflare?s massive global network, Adaptive Intelligence autonomously learns from the meta-signals of live traffic, generating short-lived rules that make automated attacks far too expensive and time-consuming to continually launch. Organizations will now have access to a powerful detection engine built to adapt and react instantly when threat actors breach their defenses. Launching an automated cyberattack used to require advanced engineering skills, but modern AI and cheap online tools have made it simple and inexpensive to carry out. Nearly anyone with a modest budget can easily rent networks of compromised devices, mask their location behind real home Internet addresses, and use free software that mimics basic human behavior. While attackers can try endless, low-cost attempts until they succeed, security teams still have to stop every threat instantly without accidentally blocking a real visitor or customer. Traditional security tools still rely on slow, scheduled updates that can take weeks or months to deploy?leaving businesses defenseless against modern threat actors who change their tactics in a matter of hours, and in some cases, minutes. Unlike conventional bot mitigation products that rely on fixed updates, Cloudflare?s Adaptive Intelligence acts like a single, constantly learning brain. By analyzing more than a trillion web visits every day to spot new threats as they happen, customers can implement Always-On Defense: Cloudflare?s ML model retrains continuously on live traffic, integrating new bot frameworks and bypass techniques into the detection engine in real time instead of waiting for manual updates or scheduled releases. Adaptive Intelligence automatically generates short-lived, hyper-targeted rules to block specific threats. Rotating rules prevents attackers from studying the system, mapping defenses, or making lasting progress. Customers can analyze multi-timeframe behavior patterns simultaneously to catch low-and-slow threats?like slow credential stuffing and scraping campaigns?that easily hide from traditional defenses. By combining browser-level session behavior signals from Cloudflare Precursor with global edge telemetry, the multi-layered architecture will be able to evaluate automation and abuse beyond basic binary tests. Security updates automatically test themselves against live traffic behind the scenes, verifying accuracy and deterring false positives before deployment with zero downtime.