TL;DR
Get privacy and security gear delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
A report on Courk’s Blog describes how its author used a custom laser fault-injection platform to bypass the RP2350’s Secure Boot protections in a security challenge and recover stored secret data. The account establishes a result on a challenge-configured Pico 2, not a general compromise of all RP2350 devices; the report also warns that its high-power infrared laser is hazardous.
A researcher writing on Courk’s Blog reports bypassing the RP2350 microcontroller’s Secure Boot with a single laser-induced fault, recovering secret data stored on a challenge-configured Pico 2. The work matters because it demonstrates a physical attack against protections designed to resist fault injection, using a custom platform the researcher describes as low-cost; it does not establish that ordinary RP2350 devices are broadly vulnerable.
The target was a Raspberry Pi Pico 2 running an RP2350 configured according to the RP2350 Hacking Challenge instructions. Those instructions put secret data in the chip’s one-time-programmable (OTP) storage, enable Secure Boot, disable debugging access and configure security features, including fault detectors. Signed firmware further restricted access to the OTP area. The challenge asked participants to recover the secret, potentially by bypassing Secure Boot.
Courk says they chose laser fault injection rather than searching for a software flaw in the Boot ROM. The report describes building a custom platform to direct a focused infrared laser at the exposed silicon die. The stated rationale was that targeting a particular area might induce a fault without triggering the chip’s voltage or electromagnetic glitch detectors. The source says a single laser-induced fault was enough to bypass Secure Boot in the challenge, but the supplied account does not provide independent verification of the result.
The report also discusses the work’s hardware and sample preparation, and says it includes manufacturing files for readers interested in building a similar platform. It warns that the setup uses a high-power infrared laser and can be hazardous. The challenge result is specific to the device configuration and attack conditions described by the researcher; the source does not show that the same method works across other chips, firmware versions or deployment settings.
A Physical Test of RP2350 Protections
The result highlights a distinction between security features that resist software attacks and protections against physical fault injection. RP2350 Secure Boot is enforced by Boot ROM code, which uses hardware support intended to make attacks harder. A fault that changes how the boot process executes could, under particular conditions, interfere with those checks and allow access that signed firmware was meant to restrict.
For people evaluating the chip, the report is a case study in how a security challenge can expose limits that may not be apparent from feature descriptions alone. Its significance should not be overstated: the account concerns a prepared board, exposed silicon and carefully timed laser pulses, rather than a remote attack or an ordinary user scenario. The researcher’s result shows a route worth examining, not a demonstrated compromise of every RP2350 product.
laser fault injection kit
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
How the DEF CON Challenge Was Set Up
Raspberry Pi introduced the RP2350 in August 2024, as an iteration on the RP2040 with additional features, including Secure Boot. A few days later, the RP2350 Hacking Challenge was launched during DEF CON 2024. The challenge used a Pico 2 and asked participants to retrieve a secret deliberately placed in OTP storage behind a locked-down configuration.
The RP2350’s Boot ROM uses a Redundancy Co-Processor (RCP) to add hardware-assisted checks. The datasheet, as quoted in Courk’s report, describes functions including stack-canary generation and validation, checking that program steps occur in the expected order, and redundant representations of values. The chip also has configurable glitch detectors that can reset the system in response to voltage or electromagnetic fault-injection attempts. For the challenge, the report says these detectors were set to maximum sensitivity.
Courk says they had experience with conventional voltage fault injection but not laser attacks, and treated the challenge as an opportunity to build a custom system. The post also points readers to other low-cost laser fault-injection efforts, including repurposed laser-engraving equipment and NetSPI’s RayV-Lite. It notes that more RayV-Lite material became publicly available by August 2025.
“I managed to beat this challenge and submitted my findings to Raspberry Pi.”
— Courk, the report’s author
Limits of the Reported Result
The source account is a researcher’s report, and the material provided here does not include independent reproduction, a public technical response from Raspberry Pi, or details confirming the precise fault’s effect across different RP2350 configurations. It is also unclear from the supplied excerpt how broadly the method applies beyond the challenge board, or what conditions would be needed to reproduce it reliably.
The report says the findings were submitted to Raspberry Pi, but the supplied material does not state whether the company validated the result, issued a security advisory, or changed any firmware or hardware guidance. The reported attack also requires physical access and die exposure; the source does not establish a remote exploit path or a practical risk to devices that have not been physically modified.
Disclosure and Reproduction Questions
The immediate next steps are to establish whether Raspberry Pi has reviewed the submitted findings and whether independent researchers can reproduce the result on other RP2350 samples. Any follow-up should clarify the affected configurations, the conditions required for the fault, and whether mitigations are available or necessary.
Readers following the work can consult Courk’s full report and the challenge materials for the author’s technical account and manufacturing files. Those files are not a recommendation to attempt the experiment: the report specifically warns that its high-power infrared laser poses a hazard. At present, the confirmed public development is the publication of the researcher’s account, while vendor validation and the scope of any wider security impact remain unsettled.
Key Questions
What did the researcher report doing?
Courk’s Blog describes using a custom laser fault-injection platform to bypass Secure Boot on a challenge-configured RP2350 in a Pico 2 and recover secret data stored in OTP memory.
Does this show that all RP2350 devices are vulnerable?
No. The report describes a result on a specific security-challenge setup. The supplied source does not establish that the method works on all RP2350 devices, firmware configurations or products.
What protections were enabled on the challenge board?
The challenge setup enabled Secure Boot, restricted access to OTP storage, disabled debugging access and enabled fault detectors. The RP2350 also uses a Redundancy Co-Processor for hardware-assisted checks in Boot ROM code.
Has Raspberry Pi confirmed the finding?
The author says the findings were submitted to Raspberry Pi. The supplied report material does not include a company statement or independent confirmation.
Is the reported setup safe to reproduce?
The report warns that its platform uses a high-power infrared laser that can be hazardous. The source does not provide a safety certification or establish that attempting the work is safe.
Source: hn
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
