'Five Eyes' intelligence alliance warns that new AI models pose urgent cyber risk
The Five Eyes intelligence alliance issued a three-page statement warning that advanced AI models will soon transform offensive and defensive cyber capabilities. Officials state that these models could bypass cybersecurity systems and breach government and business defenses within months. The alliance urges swift action to secure vulnerable software and use AI to bolster defenses.
What changed
Five Eyes agencies have issued a formal warning that frontier AI threats are months rather than years away.
Live updates
-
Five Eyes Alliance Warns Frontier AI Models Pose Urgent Cyber Risk
confidence 90%The Five Eyes intelligence alliance issued a three-page statement warning that advanced AI models will soon transform offensive and defensive cyber capabilities. Officials state that these models could bypass cybersecurity systems and breach government and business defenses within months. The alliance urges swift action to secure vulnerable software and use AI to bolster defenses.
What's confirmed:
- The Five Eyes intelligence alliance released a three-page statement regarding the risks of frontier AI models.
- The alliance warns that AI models could breach government and business defenses in months.
- Frontier AI models are expected to fundamentally transform both offensive and defensive cyber capabilities.
- The alliance states the timeline for these threats is months, not years.
Still unconfirmed:
- Anthropic's Mythos and OpenAI's GPT-5.5-Cyber allow users to execute complex and devastating hacks.
- Frontier AI poses urgent risks specifically for South Africa.
- AI puts pressure on security teams to control prompt injection and phishing risks.