Unit 42 warns AI has shifted balance of power from defenders to attackers

Palo Alto Networks' threat intelligence team, Unit 42, has warned that capabilities demonstrated by readily available agentic AI models have shifted the balance of power from defenders to attackers. Unit 42 stated that the same capabilities would be in the hands of attackers within a year. The warning specifically identifies agentic AI models as the source of the shift, describing them as readily available.

The article does not name specific incidents, but describes the current period as containing early waves of attacks that make use of agentic AI. Unit 42's projection that the same capabilities will be in the hands of attackers within a year is presented as an estimate. The team's language indicates that the balance of power has already changed, rather than merely being a future risk. The phrase 'broken in the wild' suggests that the first uses of these models in malicious activity have already occurred.

Unit 42 is the threat intelligence arm of Palo Alto Networks, a cybersecurity company. The article frames the warning as an early indicator of how agentic AI models could affect the security landscape. It does not mention any specific defensive measures, policy responses, or named threat actors. The only concrete projection is the one-year timeline for attackers to gain the same capabilities.