When stubbornness becomes a security problem: How AI agents can manipulate others
CybersecurityPsychologyComputer Science
THE AI ANGLE
Persuading other agents to alter collective network decisionsResearchers applied a social science opinion model to study multi-agent artificial intelligence networks. They showed that an unyielding agent can shift group decisions across regular communication channels, particularly from central positions. Developers and security analysts must now treat network layout and trust scoring as direct defenses against manipulation.
THE TEACHING ANGLE
Agents must remain open to new input to collaborate, yet that exact openness makes them vulnerable to manipulative peers.Read the original at techxplore.com Generate teaching or study materials
More in Cybersecurity
- Samsung and LG vow to remove 'botnet' apps from their smart TV app stores that turned sets into an AI scraping machines — but the shocking claim that over 40% of webOS apps had botnet code raises the question of how things ever got this badTechRadar · September 21, 2026
- The AI models that cheat the most, according to new CAIS benchmarkZDNet · September 21, 2026
- Should AI have the same data access restrictions as employees?CIO.com · September 21, 2026
- Presentation: The Agent Harness: Control Planes, Invariants, and Approval Boundaries for Production AI AgentsInfoQ · September 21, 2026
- Cloudflare Introduces the Agent Development Lifecycle to Replace Traditional SDLCInfoQ · September 21, 2026