Oathstar

Stories / Issue · Sep 28 – Oct 1, 2026

Rogue AI Cyberattacks Investigation

7Members · 4 D · 3 R
9Actions · 5 D · 4 R
6States & territories · CT · MO · AZ · FL lead
0Bills named

How the conversation moved

Democrat Republican · one dot = one action
Action typeFri25Sat26Sun27Mon281Tue292Wed302Thu14Total
Sent lettersTo agencies, companies, officials 1
HearingsHeld, testified or questioned 4
CriticizedA policy, decision or person 1
Filed or pushed billsNew, reintroduced or passed 3

Where they stand

No member argued against another in this window

What each party did · Democrats 5 · Republicans 4

  • 1Sent letters0
  • 2Hearings2
  • 0Criticized1
  • 2Filed or pushed bills1

By state

  • CT2
  • MO2
  • AZ2
  • FL1
  • NJ1
  • TX1

Every action

9 actions · 9 releases
  1. Oct 1, 2026
    • Bill

      DMurphy CT

      Introduced bipartisan legislation to hold AI developers criminally liable for cyberattacks by their agents.

      Release ↗

    • Hearing

      RMoody FL

      Questioned witnesses in subcommittee hearing on AI threats to children and seniors.

      Release ↗

    • Hearing

      RHawley MO

      Chaired Senate hearing on rogue AI cyberattacks on critical infrastructure.

      Release ↗

    • Hearing

      DGallego AZ

      Questioned AI experts on dangers of rogue AI models and lack of oversight.

      Release ↗

  2. Sep 30, 2026
    • Hearing

      DKim NJ

      Co-led hearing on securing homeland against rogue AI agent attacks.

      Release ↗

    • Bill

      RPfluger TX

      Introduced bill to fund AI security research and strengthen defenses against exploitation.

      Release ↗

  3. Sep 29, 2026
    • Criticism

      RHawley MO

      Criticized AI companies for recklessness in causing Hugging Face hack.

      Release ↗

    • Bill

      DGallego AZ

      Introduced resolution to establish Senate Select Committee on Artificial Intelligence.

      Release ↗

  4. Sep 28, 2026
    • Letter

      DBlumenthal CT

      Demanded answers from Treasury Secretary on AI oversight following OpenAI security breaches.

      Release ↗

Summaries, action types, amounts and positions are read from each press release by AI and can be wrong. Joint releases (same day, same action, matching title or amount) count once. How it works