Can a determined attacker capture and replay user activity to defeat Plurilock™ authentication?

Practically speaking, no. While in theory the capture and replay of user activity is possible, as a real-world task such an attack would be prohibitively complex to achieve, for multiple compounding reasons.

First, Plurilock user profiles evolve over time as users’ movement patterns evolve, meaning that captured data is only temporarily valid. Next, any capture attempt must detect low-level press, release, timing, and acceleration data, not merely the content of keystrokes or the positions of clicks, and the ability to capture this data implies an already severely compromised system.

Finally, even if somehow successfully captured and prepared for reuse in time, replay requires the replication of timing and movement patterns at a low enough level to fool the Plurilock agent(s), which likely involves a dedicated hardware attack solution—for example, simulated USB HID device(s) able to reproduce these timing and movement patterns seamlessly and faithfully.

Though in theory these things might be possible with sufficient time, skill, and resources, a successful Plurilock attack is significantly more difficult to achieve than—for example—an attack involving the theft of OTP codes delivered via SMS, which is trivial by comparison.