Meta Platforms, Inc. launched its Muse personal AI agent after CEO Mark Zuckerberg identified AI agents as a potential new source of products and revenue. Zuckerberg Sees AI Agents as Meta’s Next Growth Driver During Meta’s second-quarter earnings call in...
As Amazon blocks AI agents from shopping on its site, Walmart is taking the opposite approach, introducing new technology that lets AI agents make purchases on customers' behalf, with controls over spending and transaction approvals.
By Alexandre Kech, CEO at GLEIFAI agents can discover services, negotiate terms, analyze data, and initiate transactions for people and organizations. In financial services, this creates a governance challenge: how can other parties determine which organization stands behind an agent, who authorized it, and whether an action falls within that authority?As GLEIF’s most recent working paper on agentic payments observes, existing legal frameworks can generally allocate responsibility. The immediate
Six themes cover agent security, logging and who counts as controller, feeding a statutory code on AI and automated decisions. What will the ICO expect next?
Large Language Models (LLMs) are increasingly used for interpersonal advice, potentially shaping how users interpret ambiguous social encounters. Existing Theory of Mind evaluations typically rely on predefined answers, while uncertainty research overlooks LLMs' preferred…
Structural Causal Models (SCMs) provide a principled framework for observational, interventional, and counterfactual reasoning, all of which rely on probabilistic inference. Existing frameworks either transform SCMs into alternative tractable representations or rely on…
Scenario planning is a foresight practice in which analysts construct a small set of plausible futures and prepare an organisation for each. Each scenario is a narrative that must be plausible, internally consistent, relevant and challenging, and the set as a whole must be…
AI agents keep long-term memory, much of which comes from web pages and tool results. Agent harnesses, the software that runs the model, do not record who may act on this memory, so an instruction hidden in a fetched page can be treated as the user's own instruction. We propose…