What is reported
According to the Financial Times, Anthropic is consulting law firms to develop fair negotiation norms for AI agents. These agents would negotiate on behalf of users, for example over prices, resolve complaints or improve contract offers.
According to the report, Anthropic has already shown in experiments that stronger AI models negotiate better.
Background: a data problem
In parallel, Anthropic is hiring Conceptual Reasoning Fellows, explicitly including lawyers. For four months they would test Claude on judgment-heavy questions in law, philosophy, cognitive science, economics and political science, with pay of 100,000 to 167,000 dollars for that period.
The report points to a basic data problem: in mathematics and code, AI models are relatively easy to improve because verifiable data is plentiful or easy to generate. In the humanities such data is missing, partly because right and wrong answers are often not clear-cut.
Classification
For companies that might use AI agents in procurement or sales, the question is whether and how such negotiation norms would become binding. Read the report as an indication of a development, not as a finished rule.
Sources
Editorially owned by Ideal Syka. Sources and method: Newsroom & method. Tips and corrections: ai@i6eal.de.
