Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
A trusted execution environment is a hardware-isolated region of a processor where code and data are protected from the operating system, the hypervisor, and anyone with administrative access to the machine.

What Is a Trusted Execution Environment?
Conventional security models trust the operating system. Anything running with kernel privileges, and anyone with administrative access, can read the memory of any process. A trusted execution environment breaks that assumption deliberately: the processor enforces a boundary around a region of code and memory so that even the kernel, the hypervisor, and the physical owner of the machine cannot read inside it.
The first approach splits the processor into two worlds. ARM TrustZone divides execution into a normal world running the ordinary operating system and a secure world running a small trusted operating system, with the hardware deciding which peripherals and memory regions each may touch. Nearly every smartphone uses it for biometric templates, payment credentials, and media protection, and Apple’s Secure Enclave is a related design implemented as a separate coprocessor.
The second approach isolates at the granularity of a process or a virtual machine. Intel SGX created encrypted enclaves within a process, so the application itself was split into trusted and untrusted parts. Intel has since deprecated SGX on client processors while retaining it on server parts, and attention has moved to whole-machine isolation: AMD SEV-SNP and Intel TDX encrypt the memory of an entire virtual machine with a key the hypervisor does not hold, and ARM Confidential Compute Architecture introduces realms on the same principle. The practical attraction is that an existing workload can be lifted into protection without being rewritten.
Isolation alone would not be much use if you could not tell whether it was real. Remote attestation lets a TEE produce a signed statement describing exactly what code is running inside it, rooted in a key the hardware manufacturer provisioned. A relying party verifies that report against the vendor’s certificate chain and its own policy before releasing anything sensitive.
That is what makes the technology useful in practice. A key management service can release a key only to an enclave whose measurement matches an approved build. A data owner can send records to a processing environment run by someone else, having verified what will process them. Without attestation, a TEE is simply a claim.
It does not make the code inside it correct. An enclave with a memory corruption bug is exploitable like any other program, and the isolation then works against the defender by hiding the attack from inspection. It does not stop the host denying service, since the untrusted operating system still schedules the environment and controls its input and output. And it has a measurable history of being broken.
Foreshadow, CVE-2018-3615, used an L1 terminal fault to extract SGX attestation keys. Plundervolt, CVE-2019-11157, induced faults by manipulating processor voltage from software. SGAxe recovered sealing keys, and CacheWarp affected AMD SEV-SNP in 2023. Most of these are variants of the side-channel and speculative execution research that has run since 2018, and the pattern is that each is fixed in microcode, after which the attestation report of an unpatched platform must be rejected. That last part is a policy decision somebody has to make and keep current.
Decide what you are protecting against. A TEE answers the threat of a compromised or untrusted host operating system, hypervisor, or infrastructure operator. If that is not your concern, a trusted platform module or a hardware security module is simpler and better understood. If it is, define an attestation policy that pins the expected measurement and the minimum platform patch level, keep the trusted code small enough to review, and treat enclave code as the highest-assurance component you own.
Our Infrastructure and Architecture Review Services assess confidential computing designs, particularly the attestation policy, which is where most implementations turn out to accept more than their architecture diagram suggests.
A trusted platform module is a small fixed-function chip that stores keys and boot measurements and performs a defined set of cryptographic operations. A trusted execution environment runs your own general-purpose code inside a hardware-isolated region of the main processor, at full speed and with access to substantial memory. The TPM protects secrets; the TEE protects computation.
That is the design goal. With AMD SEV-SNP or Intel TDX, guest memory is encrypted with a key held by the processor and not available to the hypervisor, so a provider administrator cannot read it from the host. The protection depends on correct firmware, current microcode, and an attestation policy that you verify yourself rather than taking the provider’s word for it.
The main consumer use case was media protection, and the programming model required applications to be split into trusted and untrusted parts, which limited adoption. A long run of side-channel and fault injection research also undermined confidence in enclave isolation on client hardware. SGX remains supported on Xeon server processors, where confidential computing demand is concentrated.
Yes, otherwise there is nothing to distinguish a genuine protected environment from software claiming to be one. Attestation produces a hardware-signed measurement of the code running inside the environment, which a relying party checks before releasing keys or data. Any deployment that skips verification, or accepts any measurement, has the performance cost of the technology without its security benefit.
You May Be Interested In These
We use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.