Russian-speaking hackers are weaponizing Cursor AI for cyberattacks, Cisco Talos finds

3 hours ago 23

Russian-speaking threat actors have been using Cursor, the AI-powered coding environment that SpaceX recently acquired for $60B, to build malware, automate attacks, and probe for vulnerabilities, according to a report from Cisco Talos published on August 4, 2026.

The findings land at a particularly awkward moment for SpaceX, which closed its acquisition of Cursor’s parent company Anysphere just days later, around August 14.

What Cisco Talos actually found

The Cisco Talos report examined exposed AI prompt logs recovered from compromised systems. Those logs revealed that various threat actors, including Russian-speaking groups, have been feeding Cursor, alongside other AI coding assistants like Claude Code and Codex, prompts designed to generate malicious code, identify software vulnerabilities, and automate attack workflows.

Russian-speaking individuals have been running these kinds of campaigns since at least 2025, according to the report. The operational pattern involves using commercially available AI coding platforms to prototype malware, streamline credential theft operations, and accelerate the discovery of exploitable weaknesses in target systems.

The key distinction here is that Cursor isn’t being exploited in the traditional sense, where attackers find a bug in the software itself. Instead, it’s being used as intended: to write code faster and more efficiently.

Cursor’s own security issues compound the risk

Separate from the misuse-by-hackers angle, Cursor has also faced public vulnerability disclosures that raise independent concerns. Researchers have identified credential theft risks stemming from insecure configurations, as well as auto-execution flaws affecting Windows users.

Extension flaws add another layer of risk. Cursor’s ecosystem relies on third-party extensions that can introduce vulnerabilities if they aren’t rigorously vetted.

None of these vulnerabilities have been attributed to state-sponsored actors.

The SpaceX acquisition adds a new dimension

SpaceX’s $60B acquisition of Anysphere, the company behind Cursor, was one of the largest deals in the AI tooling space. Cisco Talos published its findings about Cursor’s misuse by threat actors on August 4, and the acquisition closed roughly ten days later.

Notably, there are no verified breaches involving Cursor linked to seven specific companies; available evidence reflects broader AI misuse rather than direct attribution to identified firms.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.

Read Entire Article