Cloudflare Unveils AI Defense That Makes Automated Cyberattacks More Expensive
The company said the approach is designed to prevent attackers from studying a fixed defense and repeatedly exploiting it.
Cloudflare has introduced Adaptive Intelligence, a new real-time threat detection engine designed to adapt its defenses as attackers change tactics continuously.
Built into Cloudflare Bot Management, Adaptive Intelligence uses insights from trillions of requests across Cloudflare’s global network to identify emerging attack patterns and automatically generate short-lived rules. The company said the approach is designed to prevent attackers from studying a fixed defense and repeatedly exploiting it.
The launch comes as generative AI and inexpensive online tools have lowered the barrier to launching automated cyberattacks. Attackers can increasingly rent networks of compromised devices, disguise their locations using residential IP addresses and deploy software capable of mimicking human behaviour.
Cloudflare argues that this creates an imbalance in which attackers can make repeated, inexpensive attempts, while security teams must identify and block threats in real time without disrupting legitimate users.
“Building taller walls fails when the cost of scaling an attack is effectively zero. To stop modern bot threats, you have to flip the economics on the attackers. Traditional defenses offer a static target that threat actors can systematically solve. Cloudflare’s Adaptive Intelligence creates a moving target—rendering an attacker’s engineering efforts obsolete before their operation can ever achieve scale,” said Dane Knecht, Cloudflare CTO.
Unlike conventional bot mitigation systems that depend on periodic updates, Adaptive Intelligence continuously learns from live traffic. Cloudflare said its machine-learning model can incorporate new bot frameworks and attack techniques without waiting for scheduled releases or manual rule updates.
The system also creates targeted, temporary rules that rotate over time, forcing attackers to repeatedly adapt rather than reverse-engineer a persistent security configuration.
Cloudflare said Adaptive Intelligence can analyse behavioural patterns across multiple timeframes, helping detect low-and-slow attacks such as credential stuffing and automated scraping campaigns that may evade conventional defenses.
The system combines browser-level session signals from Cloudflare Precursor with telemetry from Cloudflare’s global edge network to distinguish malicious automation from legitimate human activity.
Last month, the company launched Cloudflare OS, an open-source AI workspace designed to give every employee secure access to AI tools, internal company systems and enterprise workflows without requiring organisations to build custom AI platforms from scratch.
