Daily Briefing for 07.09.26

At a glance.

1. GhostApproval flaw exploits trust in major AI coding assistants.
2. Interpol operation cracks down on social engineering scams.
3. Chinese APT exploits Roundcube flaws to target US and Canadian universities.

GhostApproval flaw exploits trust in major AI coding assistants.

Wiz discovered a vulnerability pattern affecting six top AI coding assistants, in which a malicious repository can trick an AI agent into accessing arbitrary files on a developer's machine and potentially achieving remote code execution. The flaw, dubbed "GhostApproval," affects Amazon Q Developer, Anthropic Claude Code, Augment, Cursor, Google Antigravity, and Windsurf.

Wiz explains, "The technical primitive -- symlink following (CWE-61) – is well-known. What we found, however, goes further: in several cases, the agent's internal reasoning explicitly recognizes the dangerous target, yet the confirmation prompt shown to the user conceals this information entirely. This is CWE-451 – UI misrepresentation of critical information – layered on top of the symlink vulnerability. The user approves what they believe is a harmless local edit; the agent writes to a sensitive file outside of the project workspace."

Wiz notified all six vendors: Amazon, Cursor, and Google fixed the issue promptly, while Anthropic determined that it has sufficient measures in place to ensure that a human must confirm the file operation. Augment and Cognition (Windsurf) acknowledged receipt of the report but did not provide any further updates.

Interpol operation cracks down on social engineering scams.

An Interpol operation funded by China's Ministry of Public Security resulted in the arrests of more than 5,800 individuals across 97 countries and the seizure of $293 million, Infosecurity Magazine reports. The operation targeted perpetrators of social engineering scams, including romance fraud, business email compromise, and related money laundering. Police identified more than 15,000 suspects and blocked 31,000 bank accounts between January and April 2026.

Interpol stated, "After an initial period of intelligence collection and exchange, participating countries took part in more than three months of operational activities. This included pro-active action against high-value targets, raiding identified premises, blocking or freezing bank accounts and virtual wallets, requesting INTERPOL Notices and Diffusions and proactively utilizing INTERPOL’s Global Rapid Intervention of Payments (I-GRIP), a stop-payment mechanism that facilitates the swift blocking of illicit financial flows of both fiat and virtual assets."

Chinese APT exploits Roundcube flaws to target US and Canadian universities.

Proofpoint says a suspected Chinese APT exploited vulnerabilities in Roundcube mail servers to target physics and engineering departments at US and Canadian universities. The threat actor, tracked as "UNK_MassTraction," exploited CVE-2024-4200 to steal credential information stored in the browser, then targeted a second flaw to deploy malware on the mail server. The researchers note, "[T]he targeted departments were likely specifically chosen because they were all running versions of Roundcube vulnerable to the n-day vulnerabilities, indicating that UNK_MassTraction had conducted reconnaissance into the targets prior to conducting the campaign."

Proofpoint says the campaign primarily focused on "administrators and professors in departments with either national security ties or entities studying astrophysics and particle physics."