Taiwanese authorities have seized over NT$200 million ($6.67 million USD) in real estate and bank savings from 24-year-old Lin Rui-siang, the alleged mastermind behind the Incognito Market.

Lin, who operated under the alias "Pharoah," was arrested by the FBI at JFK Airport in May 2024 while in transit from Saint Lucia to Singapore. The seizure follows his December 2024 guilty plea in a U.S. court for narcotics conspiracy, money laundering, and selling adulterated medication.

I've been doing iOS security assessments professionally for about 15 years — banking apps, fintech, enterprise platforms. Over that time, certain patterns keep showing up in production App Store binaries. Figured it's worth sharing what I see most frequently, since many iOS developers seem genuinely unaware these issues exist.

What keeps showing up:

The most common finding is hardcoded secrets in the binary — API keys, backend URLs, authentication tokens sitting right there in plaintext strings. Developers assume compilation somehow obscures these. It doesn't. Extracting them is trivial with standard tooling.

Insecure local data storage is a close second. UserDefaults for sensitive data, unprotected Core Data databases, plist files with session tokens. On a jailbroken device (or via backup extraction on a non-jailbroken one), all of this is readable.

Weak or misconfigured encryption comes third. I regularly find apps that import CryptoKit or CommonCrypto but use ECB mode, hardcoded IVs, or derive keys from predictable inputs. The encryption is technically present but functionally useless.

Then there's the network layer: disabled ATS exceptions, certificate pinning that's implemented but trivially bypassable, and HTTP endpoints mixed with HTTPS.

Methodology:

Most of this comes from static analysis — no runtime instrumentation needed. Download the IPA, unpack, run string extraction, inspect the Mach-O binary, check plist configurations, review embedded frameworks. You'd be surprised how much is visible before you even launch the app.

I've built custom tooling for this over the years that automates the initial triage across ~47 check categories. Happy to discuss methodology or specific techniques in comments.

I've also been running a monthly live session ("iOS App Autopsy") where I walk through this process on real apps — follow the link if interested.