Researchers and practitioners are shifting from benchmark-focused AI work to formal assurance for agentic systems: on 2025-10-15 a team published a formal framework defining two models (host agent and task lifecycle) and 17 host/14 lifecycle properties expressed in temporal logic to enable verification and prevent deadlocks; on 2025-10-29 AAGATE launched as a Kubernetes-native governance platform aligned with the NIST AI Risk Management Framework (including MAESTRO threat modeling, red‐team tailoring, policy engines, and accountability hooks); control‐theoretic guardrails argue for proactive, sequential safety with experiments in automated driving and e‐commerce that reduce catastrophic outcomes while preserving performance; legal pressure intensified when Amazon sued Perplexity on 2025-11-04 over an agentic shopping tool. These developments matter for customer safety, operations, and compliance—California’s SB 53 (15‐day incident reporting) and SB 243 (annual reports from 7/1/2027) force companies to adopt formal verification, runtime governance, and legal accountability now.

Amazon’s suit against Perplexity over its Comet agentic browser crystallizes emerging legal and regulatory fault lines around autonomous AI. Amazon alleges Comet disguises automated activity to access accounts and make purchases, harming user experience and ad revenues; Perplexity says agents act under user instruction with local credential storage. Key disputes center on agent transparency, authorized use, credential handling, and platform control—raising potential CFAA, privacy, and fraud exposures. The case signals that platforms will tighten terms and enforcement, while developers of agentic tools face heightened compliance, security, and disclosure obligations. Academic safeguards (e.g., human-in-the-loop risk frameworks) are advancing, but tensions between commercial platform models and agent autonomy foreshadow wider legal battles across e‐commerce, finance, travel, and content ecosystems.

A leaked draft of the European Commission’s “Digital Omnibus” proposes major simplifications to the EU AI Act—delaying penalties until August 2, 2027, exempting some narrowly purposed systems from high‐risk registration, and phasing in AI‐generated content labels. Driven by industry lobbying, U.S. pressure, and regulatory fatigue, the draft has drawn warnings from EU lawmakers who fear weakened safeguards for democracy, rights, and safety. If adopted, the changes could shift investment and deployment timelines, complicate oversight of malicious uses, and prompt other jurisdictions to follow suit, potentially diluting global standards. Ambiguity over what counts as “high‐risk” creates a contested regulatory gray zone that may advantage incumbents and undermine AI safety and transparency ahead of the proposal’s Nov. 19, 2025 presentation.