What happens when a presidential order to halt use of a commercial artificial-intelligence tool collides with the practicalities of federal operations? Weeks after the U.S. president ordered an immediate stop to agency use of Anthropic's AI tools, federal staffers are still using the technology while agencies focus on mapping where it is embedded and finding replacements, according to reporting by Information Security Media Group (ISMG).
The immediate dilemma: an order vs. operational reality
ISMG reports that U.S. President Donald Trump ordered an immediate halt to federal use of Anthropic's AI tools. Despite that directive, federal agency staffers told ISMG they continue to use the tools weeks after the order. Officials, ISMG says, are prioritizing efforts to identify dependencies and evaluate alternatives rather than enforcing an abrupt, across-the-board cutoff.
What agencies are doing now
According to ISMG's interviews with agency staffers, the current approach centers on two tasks: first, mapping the systems, workflows and contracts that depend on Anthropic's products; second, evaluating alternative tools or paths forward. That sequence of actions — tracking usage and dependencies before switching off access — has, ISMG reports, resulted in continued day-to-day use of the technology among some federal employees.
Why officials favor tracking and evaluation over an immediate shutdown
ISMG's account indicates officials have chosen to prioritize visibility into where Anthropic's tools are embedded and to assess alternatives before enforcing an immediate stop. The reporting frames that choice as deliberate: mapping dependencies and evaluating options come first, with rapid enforcement of a shutdown taking a lower priority. ISMG's sources are agency staffers who described the situation as ongoing weeks after the president's order.
Why this matters and what to watch
- Policy implementation: ISMG's reporting highlights a gap that can emerge between a high-level directive and the operational steps agencies take to implement it — in this case, continued use while dependencies are inventoried and replacements considered.
- Agency decision-making: The reported focus on mapping and evaluation suggests agencies are treating the order as a trigger to assess impacts and alternatives, rather than as an instruction to execute an immediate, blanket shutdown.
- Timing and transparency: ISMG's timeline — noting that staffers remained users weeks after the president's instruction — underscores the temporal distance that can exist between a directive and full compliance as agencies work through practical constraints.
ISMG's reporting is the basis for this account; the organization spoke with federal staffers and described officials' prioritization of dependency mapping and alternative evaluations over rapid shutdown enforcement. The central question remains: when policy and practice diverge, who defines acceptable risk and on what timeline? The answer will shape not just how this episode concludes but how similar directives are implemented in the future.
https://www.govinfosecurity.com/federal-staffers-are-still-using-claude-despite-trump-orders-a-31437
