Valorant Anti-Cheat Bypass Methods: How Cheats Try to Beat Vanguard
A technical overview of the methods cheat developers use to try to bypass Riot Vanguard, from vulnerable driver exploitation to hypervisor techniques, and why most eventually fail.
By SensePublished
The Challenge of Bypassing Vanguard
Riot Vanguard is widely considered one of the hardest anti-cheat systems to bypass. Its boot-time kernel driver, aggressive system monitoring, and deep integration with Riot's server-side analysis create multiple detection layers that cheats need to overcome simultaneously.
Understanding how bypass methods work is valuable for both security researchers and anyone interested in the technical arms race between anti-cheat and cheat development.
Vulnerable Driver Exploitation
The most common method for gaining kernel access in Valorant cheats involves exploiting legitimate but vulnerable signed drivers. The process works like this:
Find a signed driver with a known vulnerability. Many legitimate hardware drivers have bugs that allow arbitrary memory read/write from user mode. Because these drivers are properly signed by their manufacturers, Windows allows them to load.
Load the vulnerable driver. The cheat loads this legitimate signed driver onto the system.
Exploit the vulnerability. The cheat sends specially crafted IOCTL requests to the vulnerable driver, using its vulnerability to read or write arbitrary kernel memory.
Disable or bypass Vanguard. Using kernel write access, the cheat patches Vanguard's driver in memory to disable specific detection functions, or sets up its own kernel-level infrastructure to access game memory without triggering Vanguard's monitoring.
Why It Often Fails
Vanguard maintains an extensive blocklist of known vulnerable drivers. When a driver on this list is detected, Vanguard blocks the game from launching or flags the system. The blocklist is regularly updated as new vulnerable drivers are discovered.
Additionally, Vanguard monitors driver loading at boot time. If a vulnerable driver is loaded after Vanguard, the loading event is logged and analyzed. Some cheat developers try to load the vulnerable driver before Vanguard by manipulating boot order, but Secure Boot restrictions make this increasingly difficult.
Kernel Driver Signing
Legitimate kernel drivers require a valid code signing certificate from Microsoft's WHQL program or a cross-signing authority. Cheat developers sometimes obtain these certificates through:
- Stolen certificates: Using code signing certificates stolen from legitimate companies. These certificates are eventually revoked, but the cheat may work until the revocation propagates.
- Purchased certificates: Buying certificates from companies that sell them without thorough identity verification. These are also eventually identified and revoked.
- EV certificates: Extended Validation certificates that allow kernel driver signing through the Microsoft attestation signing program.
Once a cheat driver is properly signed, it can be loaded without Secure Boot blocking it. However, Vanguard can still detect it through behavioral analysis, signature database matching, or integrity checking.
Hypervisor-Based Cheats
Hypervisor cheats operate at a privilege level below the kernel. A Type-1 hypervisor (bare-metal) or Type-2 hypervisor loads before or alongside the operating system and can intercept all memory access, system calls, and hardware interactions.
How Hypervisor Cheats Work
- The hypervisor loads early in the boot process, ideally before Vanguard's driver.
- Windows and Vanguard run as a guest inside the hypervisor without knowing it.
- The hypervisor intercepts memory access to the game process, allowing the cheat to read player positions, health, and other data without Vanguard seeing the access.
- The cheat software communicates with the hypervisor through a hidden channel (hypercall interface, shared memory region, or covert timing channel).
Why It Is Difficult
Vanguard actively checks for the presence of hypervisors. The detection methods include:
- CPUID checks: Hypervisors modify certain CPUID responses. While a well-written hypervisor can fake these responses, timing discrepancies in the CPUID instruction itself can reveal virtualization.
- Timing analysis: Certain operations take measurably longer when a hypervisor is intercepting them. Vanguard can benchmark these operations and detect the overhead.
- Hardware feature checks: Vanguard verifies that hardware virtualization features (VT-x, AMD-V) are in the expected state.
- Memory layout analysis: Hypervisors modify the system's memory layout in detectable ways, particularly around the Extended Page Tables (EPT) used for memory virtualization.
Direct Kernel Object Manipulation (DKOM)
DKOM involves modifying kernel data structures to hide a cheat from the operating system and anti-cheat. Common DKOM techniques include:
- Unlinking from process lists: Removing the cheat process from the kernel's process list so that process enumeration does not show it.
- Hiding loaded modules: Removing the cheat's driver from the loaded module list.
- Modifying callbacks: Altering kernel callback lists to prevent Vanguard from receiving notifications about the cheat's activities.
Why It Fails
Vanguard does not rely solely on standard Windows APIs to enumerate processes and drivers. It directly walks kernel memory structures and cross-references multiple data sources. If a process appears in one data structure but not another, the inconsistency itself is a detection indicator.
User-Mode Bypass Attempts
Some cheaper or less sophisticated cheats attempt to bypass Vanguard from user mode:
- Handle hijacking: Opening a handle to a trusted process and duplicating it to gain access to game memory. Vanguard monitors handle operations and blocks unauthorized access.
- Window message injection: Sending input through Windows message queues rather than direct input APIs. Vanguard validates input sources.
- Screen reading (pixel bots): Reading screen pixels to identify enemies without accessing game memory. These are hard to detect but limited in functionality.
User-mode bypasses are generally the least effective against Vanguard because the anti-cheat operates at a higher privilege level and can monitor all user-mode activity.
The Arms Race Reality
No bypass method is permanent. The relationship between Vanguard and cheat developers is a continuous cycle:
- Cheat developers find a new bypass technique.
- The bypass works for some time, from days to weeks.
- Vanguard's team identifies the technique and deploys a detection update.
- Players using the bypass receive bans, often in waves.
- Cheat developers search for the next technique.
The cost of this cycle falls on the players. Each ban means losing an account, its rank, its skins, and potentially facing a hardware ban. Even if a bypass works today, the ban may come weeks later when Vanguard deploys retroactive detection.
After the Ban
When a bypass eventually fails and a ban is issued, the hardware ban component means you cannot simply create a new account. Vanguard has already recorded your hardware fingerprint and will block any new account on the same hardware.
Sense AC's HWID spoofer handles the hardware ban side by modifying the identifiers that Vanguard reads. While no tool can undo an account ban, the spoofer allows you to start fresh on a new account with a clean hardware profile, free from the previous ban's hardware restrictions.