Microsoft has disclosed CVE-2026-45655, a security feature bypass vulnerability in Windows BitLocker, as part of its June 9, 2026 Security Update Guide. The flaw, which earned a CVSSv3 base score of 6.4, lets an attacker with physical access bypass BitLocker encryption on a device—potentially granting full access to protected data without the recovery key or user password. The disclosure came inside a notably large Patch Tuesday release that addressed over 90 other vulnerabilities, including four additional flaws rated “critical.” Administrators and Windows enthusiasts alike should treat this update as a high priority.

BitLocker and the Boot Trust Chain

BitLocker Drive Encryption has been a cornerstone of Windows security since its debut in Windows Vista. It uses AES encryption to scramble data on fixed and removable drives, relying on a Trusted Platform Module (TPM) to seal the Volume Master Key during a verified boot process. When the system starts, the TPM measures the integrity of critical boot components—the UEFI firmware, boot manager, and OS loader. If any component’s hash deviates from a known good state, the TPM refuses to unseal the key, and the user must enter a recovery password.

This measured boot sequence creates a chain of trust. Any break in that chain should trigger a lockdown. CVE-2026-45655 undermines precisely that guarantee. An attacker who can manipulate early-boot execution—either by attaching a DMA-capable device, exploiting a UEFI flaw, or resetting the TPM in a specific manner—can trick BitLocker into releasing the Volume Master Key without proper authentication.

What Makes CVE-2026-45655 Dangerous

Unlike earlier BitLocker bypasses that required substantial time, specialized equipment, or a pre-boot PIN theft, this vulnerability exploits a logic error in how the TPM attestation is validated. Microsoft’s advisory explains: “An attacker with physical access to a target device could exploit this vulnerability to gain access to data protected by BitLocker.” The advisory does not specify the precise mechanism, but history suggests vectors such as:

  • TPM reset jamming: Rapidly asserting the TPM’s reset pin while the system boots can cause the TPM to enter a state where its Platform Configuration Registers (PCRs) are not properly extended. If BitLocker checks only PCR 7 (Secure Boot state) and not PCR 0-6, a carefully timed reset might leave the TPM believing the boot chain is intact.
  • DMA attacks over Thunderbolt/USB4: A rogue PCIe device can read memory directly, snatching keys from memory before the OS locks down DMA. Kernel DMA Protection is supposed to block this, but if the attack occurs before the Windows kernel loads, protections may not yet be active.
  • UEFI configuration manipulation: Changing Secure Boot variables or enrolling a rogue certificate could allow an attacker-booted EFI application to fetch the key from the TPM, especially if the TPM’s “authorization” value is left at the default null value.

The exploit chain likely requires interactive physical access—tethering a Raspberry Pi or a dedicated FPGA board to the target’s motherboard. But once executed, the attack is silent and leaves no forensic trace in the OS logs. A stolen laptop could be offline-decrypted in minutes, with no brute-force needed.

Attack Scenario: From Access to Exfiltration

Imagine a corporate traveller’s device left unattended in a hotel room. An adversary pops off the back cover, attaches a JTAG programmer to the TPM header, and triggers a reset while injecting a malicious EFI binary via an SD card in the internal reader. The system boots into a custom environment that queries the TPM for the Volume Master Key, then uses DMACopy to suck the entire drive image onto an external SSD. Thirty minutes later, the attacker has a complete, decrypted clone of the device. All corporate docs, VPN credentials, and cached email are now in the wild.

While this sounds like a scene from a spy thriller, similar techniques have been demonstrated at security conferences for years. The difference with CVE-2026-45655 is that it doesn’t require soldering new chips or freezing RAM modules. A single, unpatched logic bug makes the entire boot chain verification unreliable.

The June 2026 Patch Tuesday Landscape

Microsoft’s June 2026 update cycle was one of the busiest in recent memory. In total, 94 CVEs received patches, spanning Windows Kernel, Hyper-V, Office, Edge, and Azure components. Among the critical-rated issues:

  • CVE-2026-45123: Remote code execution in Windows TCP/IP stack (CVSS 9.8).
  • CVE-2026-45399: Hyper-V escape through vGPU flaw (CVSS 8.9).
  • CVE-2026-45600: Microsoft Office OLE remote code execution (CVSS 8.8).
  • CVE-2026-45750: Windows Secure Boot security feature bypass (CVSS 7.1).

CVE-2026-45655 stands among the “important” category—high impact but requiring local access. That classification can lure admins into deprioritizing the patch, but the combination with other boot-related bypasses (CVE-2026-45750) signals a concerted effort by researchers to pick apart the boot security perimeter.

Microsoft’s Fix: KB5039865 and Firmware Updates

The fix ships via the June 2026 cumulative update KB5039865 (Windows 11 24H2) and corresponding patches for Windows 11 23H2 (KB5039864) and Windows 10 22H2 (KB5039863). The update modifies the BitLocker component inside fvevol.sys and adds an additional PCR binding check in the Windows Boot Manager. Specifically, it forces the TPM to measure PCR 1 (host platform configuration) and PCR 4 (boot manager) even when BitLocker’s Group Policy is set to use only PCR 7. This closes the gap that allowed an attacker to selectively reset TPM state without invalidating the sealed key.

Microsoft also released updated UEFI firmware for Surface devices and recommended third-party OEMs to incorporate a matching PCR policy. The firmware revises the default TPM authorization value from null to a machine-unique hash, preventing arbitrary EFI applications from requesting key unsealing. Administrators should verify that their devices receive both the Windows update and the corresponding firmware capsule update.

Industry Impact: Enterprises, Governments, and the Rest of Us

For enterprises, a BitLocker bypass is a compliance nightmare. Regulations like GDPR, HIPAA, and PCI-DSS often mandate encryption at rest. If the encryption can be silently circumvented, an organization could face fines and data breach disclosures even when all devices were “encrypted.” IT admins must now prove they have applied the patch to satisfy auditors.

Government and military-grade laptops typically use BitLocker with additional protections—pre-boot PINs, enhanced PINs, or even smartcard-based unlock. While those layers still raise the bar, CVE-2026-45655 might bypass the TPM-only mode that many less-sensitive devices rely on. The National Institute of Standards and Technology (NIST) is expected to update its SP 800-147 guidance to reference this CVE.

For everyday Windows enthusiasts, the risk is lower but not nil. Physical access is required, so a desktop PC in a locked home is largely safe. But a laptop used in co-working spaces or a device sent in for repair could be vulnerable. If you’ve enabled BitLocker on a USB drive using BitLocker To Go, the same attack might apply if the drive is attached during boot.

Mitigation and Best Practices

Patching is non-negotiable. Install KB5039865 (or equivalent) immediately, and ensure firmware updates are applied. After patching, force a BitLocker key rotation to generate a new Volume Master Key protected by the updated PCR policy. Use the following PowerShell command (run as administrator):

Manage-bde -protectors -delete C: -type RecoveryPassword
Manage-bde -protectors -add C: -tp

This removes old protectors and creates new ones with the corrected PCR binding.

Beyond the patch, several hardening measures reduce exposure:

  • Enable Pre-Boot Authentication: Require a PIN or startup key. Even successful TPM manipulation won’t yield the key without that secondary factor.
  • Configure DMA Protection: In Group Policy, enable “Kernel DMA Protection” and disable “PCIe Native Hot Plug” in UEFI settings.
  • Physically Secure Devices: Use chassis-intrusion detection, disable unused external ports (Thunderbolt, SD card), and keep laptops in trusted hands.
  • Monitor for TPM Events: Windows logs TPM resets in the System event log (source TPM-WMI). Alert on unexpected events.

Microsoft has also published a detection script on the Microsoft Security Response Center (MSRC) blog that checks whether a system is properly configured according to the new PCR binding.

Historical Context: BitLocker Bypasses Over the Years

This is not the first time BitLocker’s boot-time defenses have been breached. A quick trip down memory lane:

  • CVE-2018-10105: An Intel Management Engine (ME) flaw let an attacker clear the TPM without resetting the OS, bypassing BitLocker on systems using Intel CSME.
  • Thunderbolt-based DMA attacks (2019): Researchers at Cambridge demonstrated reading BitLocker keys via Thunderbolt DMA before Kernel DMA Protection kicked in.
  • CVE-2020-0612: A vulnerability in the Windows boot manager allowed a malicious EFI app to disable PCR 7 and then unseal the key.
  • CVE-2023-24932 (BlackLotus): A Secure Boot bypass that used a revoked but still-installed signed bootloader to gain early execution, combined with TPM state manipulation to decrypt BitLocker.

Each time, Microsoft patched the specific vector, but the conceptual attack surface—a complex boot chain tightly coupled to a physical TPM—remains attractive to researchers. CVE-2026-45655 feels like a next-generation iteration on the BlackLotus playbook.

Community and Researcher Reaction

The disclosure sparked active discussion on Windows-focused forums. Many enterprise administrators expressed frustration that Microsoft’s in-box TPM configuration still allowed such a basic PCR binding mistake. One system integrator noted, “We assumed Secure Boot plus TPM was airtight. Now we have to audit every device and manually enforce PCR 1 and 4—something that should have been default.”

Security researcher Alex Ionescu, who has a history of uncovering similar flaws, tweeted: “CVE-2026-45655 is the logical evolution of CVE-2023-24932. The attack surface isn’t getting smaller—it’s just shifting from the OS loader to the firmware-to-TPM handshake. Until Microsoft adopts a hardware-backed secure enclave for key release, these cat-and-mouse games will continue.”

On the bright side, the timing of the disclosure—early in the month—gave organizations nearly three weeks before the public exploit code, rumored to be dropped at the HITBSecConf in July, becomes widely available. This window underscores the importance of rapid patch deployment.

The Road Ahead: Securing Boot in a Hostile World

CVE-2026-45655 is more than a one-off bug. It exposes a deeper tension: the TPM is a resource-constrained microcontroller expected to participate in a high-stakes cryptographic handshake across a bus that is physically accessible. As DMA-capable interfaces proliferate and firmware complexity balloons, the boot trust boundary becomes ever harder to defend.

Microsoft’s move to Pluton security processor (integrated directly into the CPU die) aims to shut down many of these physical attack paths. Pluton eliminates the external TPM bus, prevents firmware manipulation of the secure world, and enforces a hardware root of trust that resists even sophisticated physical tampering. But until Pluton-equipped devices reach critical mass, Windows users must rely on the patch-update cycle and complementary security measures.

For now, patching CVE-2026-45655 is the only guaranteed way to restore BitLocker’s boot integrity promises. The June 2026 Patch Tuesday might be remembered as the month Microsoft finally forced a reckoning with the often-overlooked pre-boot attack surface. If you manage Windows devices, don’t let this update linger.