How to constrain an AI agent properly

How to constrain an AI agent properly

Main BTU researchers’ assessment is that useful constraints should become stronger where an AI agent’s actions have wider consequences or

Who sets the boundaries for AI agents

Who sets the boundaries for AI agents

Main BTU researchers’ assessment is that an AI agent’s boundaries become stronger when accountability matches decision authority: the human process

One AI agent or several

One AI agent or several

Main BTU researchers’ analytical assessment is that an additional AI agent creates value when a task needs a specific missing

Who Will AI Agents Empower?

Who Will AI Agents Empower?

Main BTU researchers’ assessment is that an AI agent can create a new opportunity for a person, but it can

How an AI agent team works

How an AI agent team works

Main BTU researchers’ analytical assessment is that an AI agent team becomes dependable when evidence, uncertainty and the exact state

What an AI agent must never do

What an AI agent must never do

Main BTU researchers’ analytical assessment identifies a central condition for safe organizational agents: an agent must not be able to

What permissions should an AI agent have

What permissions should an AI agent have

Main BTU researchers’ analytical assessment is that useful AI autonomy depends on precisely chosen permissions sufficient to complete a defined

Why an AI Agent Needs a Mandate

Why an AI Agent Needs a Mandate

Main BTU researchers’ analytical assessment is that an AI agent gains useful organizational independence when its principal, purpose and boundaries

Why Not Every Task Needs an AI Agent

Why Not Every Task Needs an AI Agent

Key Takeaway As AI agents become more capable, organizations face an understandable temptation: if a system can plan, use tools