SIIA Statement on Anthropic and Department of War Negotiations

The following statement can be attributed to Chris Mohr, President, the Software & Information Industry Association (SIIA).

In order for AI to be successfully deployed in a democratic society, it must be adopted with appropriate risk-based guardrails. We support Anthropic's decision to work with the Department of War (DoW) to deploy its AI models to advance national security while also requesting reasonable limitations on the use of those models in a narrow set of cases. We share Anthropic's view that mass domestic surveillance is incompatible with democratic values. We also agree that fully autonomous weapons require AI systems that are suited to the task - requiring a degree of reliability that Anthropic acknowledges has not yet been achieved. Very few DoW use cases even touch on these situations.

We encourage the parties to find agreement and caution against counterproductive measures. Invoking the Defense Production Act to compel the removal of security restrictions, or designating a domestic leader like Anthropic as a 'supply chain risk,' represents an overbroad response to a technical disagreement. Such a 'blacklisting' approach, typically reserved for hostile foreign entities, is both untethered from the facts of Anthropic's security posture and unlikely to advance a long-term solution.