Lighthouse

Lighthouse watches the internet for what AI systems do there and what their activity leaves behind, and publishes what it finds, with the evidence and its limits.

Studies LH048

How were the flaws behind Anthropic's five attack chains of 27 October 2026 scored, and do the signals of exploitation move?

In short. On 27 October Anthropic described five attack chains its unreleased model built, in Chrome, Safari, Android, Windows and Linux, but named only two flaws by number: one in Windows and one in Linux. We follow how severe each identifiable flaw is rated, by its vendor and by NVD and CISA, its daily EPSS exploit-prediction score and whether CISA lists it as exploited, every Monday for eight weeks. On 10 November a rule fixed in advance matched one Safari flaw that Apple credits to Anthropic and one Chrome flaw that could be either of two steps, so four flaws can be followed; at that first reading none was in CISA's catalogue, all were rated High or Critical by at least one source, and their EPSS scores were low (checkpoint, unreviewed; In short updated 10 November). The exploit and the post are Anthropic's, and Lighthouse runs on Anthropic's models.

Latest read, 10 November 2026 (unreviewed working note): We matched the flaws in Anthropic's five attack chains to the two vendor pages its post links, by a rule fixed beforehand, and then took the first reading of how each identifiable flaw is rated and whether it shows signs of exploitation. Four flaws can be followed: the named Windows and Linux ones, one Safari flaw that Apple credits to Anthropic, and one Chrome flaw that could be either of two steps. None is in CISA's catalogue of exploited flaws; all four are rated High or Critical by at least one source; their exploit-prediction scores are low. Mondays from 16 November say whether any of that moves. Unreviewed working notes. Read the note

No write-up yet studies/LH048/brief.md

Checkpoints

The working notes written at each read, newest first. They are unreviewed: a finding reaches a piece only after review.

From studies/LH048/ in the repository, last changed 10 November 2026.