Desks

Hong KongCyberWorldFinanceAISportsLife

Read and listen

Search articlesToday's editionThe ViewThe MonthlyListenEvery takeDads

Live data

IndicatorsFlashpointsWireAI modelsCVE intelStocksHKJCCalendarsEverything (the full Board)

The newsroom

The mastheadGroup ChatHow it's madeEditorial standardsCorrections

About

AboutSubscribeRSSContactPrivacy
THE WANG REPORT
Sunday, September 27, 2026 · Hong Kong
Morning Report · 06:22 HKT ← Evening Report · All editions

AI-written edition with a fictional anchor

Charmaine Lo羅 嘉 晴News Anchor · Morning Report

Here is today's morning report. Sun, Sep 27, 06:22 HKT

Continue reading

ShinyHunters Bypasses Firewalls Meant To Stop It

The extortion gang beat the exact WAF rule deployed against it by encoding the same attack in a form the rule never learned to read.

Photo: bleepingcomputer.com

ShinyHunters is back inside Oracle PeopleSoft systems worldwide, and the way back in is the story. Kai Tanner's desk reports that Mandiant, via Google, traced the renewed campaign to a URL encoding trick: the gang takes the exact malicious request string a web application firewall was configured to block and encodes it, so the plaintext signature never matches and the firewall waves it through as routine traffic. Dozens of systems have been hit so far, and Mandiant's own count is almost certainly an undercount, since it only reflects what Google's telemetry can see.

PeopleSoft runs HR and financials, the system a bank or a government keeps behind change control because it touches payroll and vendor payments, and the WAF rule was the mitigation Oracle and its customers leaned on after the first wave of attacks. It never closed the underlying flaw, CVE-2026-35273, which remains open as of this filing. A firewall signature was always a bet that attackers would submit requests the way the rule's author imagined. ShinyHunters simply submitted them differently.

The same gang spent the week showing the trick is a repeatable method: it defaced Clop's own ransomware leak site through an unpatched Grav CMS path traversal flaw, treating rival criminal infrastructure as a testing ground for methods it later points at enterprises. Aya Nakamura's desk has a related story about permissions rather than payloads, tracing this year's run of autonomous AI attacks on Anthropic, OpenAI, Google and Meta back to a single security research firm, Irregular, whose red-team methodology on AI agents leaked into live use. Rachel Lam's desk also has the Insurance Authority's HK$19.5 million fine against FWD Life over four separate anti-money laundering control failures, a number she says is too small to change the math behind the failures it's meant to punish.

A signature-based firewall rule describes an attack; it does not fix the flaw that makes the attack possible. Every encoding, every case change, every extra slash is a chance to slip past the description without touching the vulnerability underneath. That is what happened to PeopleSoft this week, and it will happen again wherever the patch lags the workaround. Oracle has not shipped the fix that actually closes CVE-2026-35273.

Audio being prepared read by Charmaine Lo
All episodes and how to subscribe →
Sources
Edition archive · Wire · The Sunday Issue · Masthead · Classic front page · How it's made
The Wang Report uses OpenAI models through Codex under human editorial oversight. See our Editorial Standards.