Brian Armstrong: New AI controls make crypto finance safer
Coinbase CEO Brian Armstrong said on X that new AI controls will make using AI in financial accounts safe, urging users to embrace change.
AI safety refers to technical and governance measures that reduce risks from artificial intelligence systems. In crypto news it appears in Coinbase's rollout of AI safeguards, OpenAI's rapid shutdown system, and academic findings on AI agents sharing workspace data.
Coinbase CEO Brian Armstrong said on X that new AI controls will make using AI in financial accounts safe, urging users to embrace change.
Researchers found that GPT‑6 agents can retrieve each other's intermediate results from a shared scratch store, achieving near‑perfect success on paired arithmetic problems.
OpenAI introduced a centralized safety system that can stop research model workloads in under half a minute, aiming to curb rogue behavior.
It is the set of tools and policies that aim to prevent AI systems from causing financial loss or security breaches in crypto platforms.
Coinbase is integrating AI controls that monitor and protect user accounts, a move described by CEO Brian Armstrong as essential for competitiveness.
OpenAI's new Control Room can revoke a model's output authorization in as little as 35 seconds, allowing rapid response to rogue behavior.