Understand what happened
Google Project Zero published a technical research post describing a new way to bypass Address Space Layout Randomization (ASLR) on Apple devices by abusing how pointer-keyed collections (like NSDictionary) are serialized and re-serialized.
The method can leak in-memory pointers in scenarios where a service accepts attacker-controlled serialized data and then returns a re-serialized object — a pattern that exists in a surprising number of real-world APIs and daemons. Project Zero published the write-up on Sep 28, 2025.
Important Google Project Zero update:
This is important because ASLR is a core mitigation that makes exploitation of memory bugs harder. A reliable pointer leak can turn a theoretical memory bug into a remotely exploitable chain.
Project Zero’s disclosure does not publish exploit code for attacks in the wild, but it shows a practical technique researchers and vendors must consider when hardening serializers and IPC services.
What Project Zero found and why the ASLR bypass matters
Project Zero’s research shows that certain pointer-keyed data structures can leak internal pointer values when the system serializes them to a transferable format and later re-serializes them in a way that reveals bucket placements or sentinel values.
In practice, an attacker who can send crafted serialized data to a vulnerable service and receive the re-serialized response could infer memory addresses — effectively removing ASLR’s randomness for targeted processes.
Why that matters

ASLR stops attackers from knowing where libraries, stacks, and other memory regions live. Once ASLR is bypassed, classic memory-corruption exploits (use-after-free, ROP chains) become far easier to build and deploy.
Even without a memory corruption bug today, an information leak like this raises the risk profile for any system that deserializes untrusted input.
Which Apple components are affected and who should worry
Project Zero’s write-up explains the bypass in general terms and shows how it applies to object serialization patterns commonly used across iOS and macOS services and apps. The post does not name a specific CVE tied to widespread active exploitation, but it flags a class of behaviors developers and Apple’s security teams should harden.
Apple often waits to publish patches after coordinated disclosure; users should expect vendor advisories and updates if Apple finds code paths that match the pattern.
Short version for users: if you use iPhone, iPad or Mac, this research is a reason to keep your devices updated and avoid installing untrusted apps that expose network/IPC surfaces. Enterprises should accelerate patch testing when Apple issues fixes and review any internal services that deserialize external data.
A high-level technical view — safe to read, no exploit steps
Project Zero’s technique is notable because it finds a non-memory-corruption path to leak internal pointers. In plain terms:
- Some collections place items into internal “buckets.”
- The process of serializing and then re-serializing those collections can reveal how items were arranged.
- Arrangement depends on pointer values; observing arrangements across crafted inputs leaks pointer bits.
That leak is enough, in many cases, to defeat ASLR without resorting to timing side-channels or classic memory corruption. I’m intentionally avoiding step-by-step guidance — the responsible approach for public security reporting — but Project Zero’s post goes into the research details for vendors and defenders.
Vendor, developer and defender actions

Project Zero recommends several mitigation paths that developers and platform maintainers should pursue:
- Treat deserialization of attacker-controlled content as risky and add strict validation.
- Avoid returning raw re-serialized objects that contain pointer-dependent structure to untrusted callers.
- Apply additional runtime hardening (heap/isolation techniques) and enforce least privilege for services that accept serialized payloads.
For Apple and large vendors, the practical fixes may include API redesigns, server-side filtering of serialized inputs, or turning off automatic re-serialization for exposed IPC endpoints. Security teams should also run targeted audits to find other serializers that match the risky pattern.
What to watch next: patches, advisories and CVEs
- Apple advisories: Apple typically issues security updates after investigation; check Apple’s security update pages and apply OS patches promptly.
- Vendor responses: Developers of apps and frameworks that perform serialization should review Project Zero’s write-up and patch serialization logic where needed.
- Regulatory/enterprise: Enterprises with managed fleets should prioritize patch testing and do risk assessments for services exposed to untrusted inputs.
Frequently Asked Questions
What is the ASLR bypass on Apple devices?
Google Project Zero disclosed a method that can leak memory addresses via `NSDictionary`-style serialization and re-serialization, allowing attackers to defeat ASLR in some scenarios. The technique works when a service deserializes attacker data and later returns a re-serialized object.
Does this mean iPhones and Macs are immediately exploitable?
Not necessarily. The research shows a technique, not a public exploit chain in the wild. However, the method raises risk because it can make other memory bugs easier to exploit. Users should apply updates when Apple issues patches.
Should I stop using my Apple device right now?
No. Take practical steps: keep software updated, avoid untrusted apps and attachments, and follow guidance from Apple or your IT team. The main defensive action is timely patching and reducing untrusted deserialization surfaces.
What can developers do to mitigate the risk?
Developers should validate or refuse untrusted serialized inputs, avoid returning raw re-serialized objects to callers, and apply runtime hardening to services that handle external data.
Where can I read Project Zero’s research?
Project Zero published the research on its blog with examples and defensive notes for vendors. Security teams should read the original post for technical context.
Author note: I’m a security reporter summarizing Project Zero’s research and coverage from security outlets. I intentionally omitted exploit code or step-by-step instructions to avoid enabling malicious use. The article focuses on defensive steps for users, developers and IT teams. Read the Project Zero blog for technical depth and check Apple security updates for vendor patches.




