▲Autonomous Agents Move Into Cyber Offense

Score history

Daily conviction score, 0 to 100. Higher means the thesis is more strongly corroborated.

Now 93 · +2 since Oct 9 · ranged 91 to 93

Showing the last few days. Unlock full score history.

Why the conviction moved

• Oct 10
  Strengthened +2

  The open-source REA tool gives coding agents reverse-engineering and decompiler access through the Model Context Protocol, lowering the cost of agent-driven vulnerability research. Sophos's report that agents now close half of its managed detection cases shows defenders automating too, which tempers the 'outpacing defenses' half of the thesis.

• Oct 9
  Strengthened +3

  A second paper finds that AI triage agents miss 40% of attack alerts and dismiss them when their own searches return no records. Automated defenders are failing in ways attackers can exploit, which widens the offense-defense gap.

Showing the last 2 days. Unlock the full record.

Source trail

• Supporting · October 10, 2026

  Windows 11 ships enforced sandboxes for AI agents as an open-source tool hands agents a decompiler

  The open-source REA tool gives coding agents reverse-engineering and decompiler access through the Model Context Protocol, lowering the cost of agent-driven vulnerability research. Sophos's report that agents now close half of its managed detection cases shows defenders automating too, which tempers the 'outpacing defenses' half of the thesis.

  AI ML Big Data

• Supporting · October 9, 2026

  Two Security Papers Find Agent Defenses Miss Attacks: Skill Scanners Bypassed 94% to 100% of the Time, Triage Agents Miss 40% of Alerts

  A second paper finds that AI triage agents miss 40% of attack alerts and dismiss them when their own searches return no records. Automated defenders are failing in ways attackers can exploit, which widens the offense-defense gap.

  arXiv cs.CR