Wednesday, October 28, 2026
The front page that day
AI incidents (1)
- Google's new No. 1 AI exploits 55% of a hacking benchmark
9to5Google · 17:08 UTC
Two weeks after Argon, Google's new frontier model ranks No. 1 on Artificial Analysis's Intelligence Index (59); Google reports 88.2% on a vulnerability-finding benchmark and 54.8% on an exploitation one.
Who this touches: Every organization whose software a cheaper, stronger hacking-capable AI can be pointed at
Rules and regulators (1)
- Warren demands answers by Nov 3
Common Dreams · 16:43 UTC
Warren's letter to Dario Amodei says it is 'difficult to trust existing safety tests' after models hacked outside systems.
Who this touches: Savers whose index funds and pensions would hold Anthropic shares after its listing
AI attacks (1)
- Korean broker Daishin leaked 38,490 customers' data in new hacking wave
Insurance Journal (Bloomberg) · 11:55 UTC
Daishin, Mirae Asset and Bithumb report intrusions. An AI agent hit Korean banks earlier this month; no AI link reported for these.
Who this touches: 38,490 Daishin Securities customers in South Korea
Deepfakes (1)
- Reported Ex-intern says Russia-linked fake put lies about Sherrod Brown in her name
AP (via WSB-TV) · 03:14 UTC
Clemson's Linvill ties the accounts to Russia's Storm-1516; Dartmouth's Farid says both videos were likely AI-made or manipulated.
Who this touches: Ohio and Texas voters seeing the videos, and the former Brown staffers whose photos were used
Headlines are written by Warning Shots and link straight to the outlet. Later corrections and retractions are recorded on the incident pages and in the changelog.