Cybersecurity and national security leaders should stop treating artificial general intelligence as a distant thought experiment and start planning for a range of near-term AI scenarios, a new RAND analysis warns. The report says forecasts across expert surveys, prediction markets and compute-based models have all shifted earlier, with some central estimates now landing in the 2030s.
That matters for cyber defenders because RAND argues dangerous capability jumps could reshape the attacker-defender balance before anyone agrees AGI has arrived. At the same time, it says the forecasting system itself is immature, lacking strong benchmarks, real-time capability tracking and enough independent validation.
The practical takeaway is simple: plan around scenarios, not a single date. RAND says agencies should set reassessment triggers and watch closely for signs that AI is accelerating AI research itself.
Image by Gerd Altmann from Pixabay
