ASSESSMENT
OSC performed a review of all applications known to be in use by the agency to determine if covered AI is utilized.
DETERMINATION
Upon performing the assessment, the CAIO has determined that OSC does not use and does not anticipate using covered AI.
REASSESSMENT
OSC will reassess its enterprise environment every six months for active AI uses.
***
OSC Policy on AI
The Office of Special Counsel (OSC) establishes these principles and policies for the use of AI as it relates to OSC's mission, and in support of the President's Executive Order (EO) 14110 on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence, Office of Management and Budget (OMB) Memorandum M-24-10, and the Advancing American AI Act of 2022.
OSC's mission is to safeguard the merit system by protecting federal employees and applicants from prohibited personnel practices (PPPs), especially reprisal for whistleblowing; operate as a secure channel for federal employees to blow the whistle and ensure that the disclosed wrongdoing is investigated and, where necessary, corrected; enforce the Hatch Act, which puts certain restrictions on partisan political activity by government employees; and protect the civilian employment and reemployment rights of military service members under the Uniformed Services Employment and Reemployment Rights Act. Because AI could impact both the agencies over which OSC exercises jurisdiction, as well as the internal operations of OSC itself, the policies stipulated below serve as a guide for OSC's approach in both contexts.
Guiding Principles
- Review of AI Usage – OSC is committed to ensuring that, when AI use is implicated in a claim within OSC's jurisdiction, it is used lawfully and properly in the federal government.
- Efficiency and Effectiveness – insofar as AI tools can assist OSC's mission in a safe, responsible, and efficiency-enhancing way, the agency will explore ways to deploy them within our day-to-day work.
- Transparency – OSC's use of AI will follow proper vetting and compliance with federal requirements for security and privacy; it also will be reported appropriately to OMB and the public.
Oversight of the Use of AI
As the use of AI proliferates throughout government agencies over which OSC exercises jurisdiction, OSC is ready to receive and evaluate whistleblower disclosures related to Executive Branch usage of AI that may evidence violations of law, rule, or regulation; gross mismanagement; a gross waste of funds; an abuse of authority; a substantial and specific danger to public health or safety; or censorship related to scientific research. To this end, OSC expects to be knowledgeable and aware of how federal agencies use AI tools in various contexts, such as, for example, hiring, disciplinary actions, or performance review decisions.
OSC's Use of AI
OSC does not currently have use cases for “covered AI" as defined by OMB Memorandum M-24-10. OSC stands ready to embrace new technologies, including AI, if they are an efficient and effective means to further OSC's mission. When OSC identifies use cases for covered AI, a consolidated, machine-readable file of all publicly releasable use cases will be posted and maintained at https://osc.gov/ai and sent to OMB, as required by the memorandum.
As we explore the possible use of these AI technologies, data security and privacy are of the utmost importance. Prior to the deployment of any AI system into an OSC workflow, our technical professionals will thoroughly vet the technology to understand its behavior and to ensure the integrity of confidential information and will advise and train staff on its proper use. OSC places a particular emphasis on protecting the personal information of individuals who file complaints with OSC and the contents of OSC's ongoing investigations, as required by federal privacy regulations.
OSC commits that any future use of AI technology by the agency will be efficiency-enhancing, compliant with existing federal requirements for security and privacy, and conformant with the rules outlined in EO 14110 and OMB M-24-10.