Top Stories – September 09, 2026
1. EU AI Act: Frontier GPAI Systemic Risk Evaluations Due September 15
Source: Cubbbix / AI Regulation September 2026 | Date: September 9, 2026
Providers of GPAI models trained above the 10^25 FLOPs threshold must submit their first formal systemic risk evaluations to the European AI Office by September 15, covering red-teaming results, energy consumption disclosures, and compliance with the EU copyright training-data summary template.
2. Left-Right Coalition Presses White House to Release Classified AI Model Review Framework
Source: Semafor | Date: September 9, 2026
More than 25 advocacy groups spanning the political spectrum sent a letter today demanding the White House publicly release the EO 14409 frontier model pre-release review criteria; Protect Democracy separately filed a FOIA lawsuit September 1 against the White House and Commerce Department after the administration declined to disclose the classified benchmarking standards.
3. NIST Releases Draft AI Cybersecurity Quick-Start Guide for CSF 2.0 Compliance (SP 1353)
Source: NIST | Date: August 19, 2026
NIST published an initial public draft of SP 1353, providing structured AI prompt templates and three use cases for Cybersecurity Framework 2.0 compliance analysis and reporting. Public comments are open through October 15, 2026.