The U.S. government has cleared a major barrier in expanding artificial intelligence across civilian operations. OpenAI, Google, and Anthropic have been approved as official AI vendors, simplifying access to their tools for federal agencies. This move eliminates long procurement processes and positions these companies at the centre of the government’s AI infrastructure. Federal departments can now integrate models like ChatGPT, Gemini, and Claude more swiftly and securely.
GSA Fast-Tracks AI Adoption Across U.S. Federal Agencies
The General Services Administration (GSA) announced the addition of OpenAI, Google, and Anthropic to its Multiple Award Schedule, a key contracting mechanism that pre-negotiates prices and terms for federal purchases. This development allows agencies to bypass individual contract negotiations and instead access vetted AI tools directly. GSA officials said the decision is not about favoring particular firms but enabling a broader rollout of AI.
“We’re not in the position of picking winners or losers here,” GSA Deputy Administrator Stephen Ehikian said. “There’s going to be different tools for different use cases.” He emphasized the agency’s goal of providing a wide range of AI tools to boost productivity across departments.Previously, each agency had to navigate its own legal and procurement hurdles to adopt AI tools. This slowed adoption and limited most AI projects to small-scale pilots or national security applications. Now, with three major vendors in place, large-scale and cross-agency AI implementation can begin.
Rigorous Vetting Ensured AI Safety, Performance, and Bias Checks
The inclusion of OpenAI, Google, and Anthropic followed thorough evaluations by the GSA. Officials stated that each vendor’s language models underwent detailed assessments for safety, security, performance, and bias. These tests were necessary to meet federal standards and ensure the tools align with evolving national AI policy.Despite political tensions around perceived ideological bias in commercial AI models, the GSA confirmed that the selection focused on utility and capability.
Read also:U.S. Leads the Global AI Race
“The models were vetted using several performance and security measures,” said Josh Gruenbaum, Commissioner of the GSA’s Federal Acquisition Service.While some critics voiced concern about the inclusion of companies accused of liberal bias, GSA officials reiterated that the evaluations prioritized technical benchmarks. They added that additional AI vendors could join the list once they complete the same review process.
Agencies Plan to Deploy AI for Chatbots, Fraud Detection, and Policy Review
A number of such agencies have already put plans on how they will utilize the newly approved AI tools. The Office of Personnel Management (OPM) will implement AI in customer support chatbots and to understand the reaction of citizens on suggested regulations. According to OPM Director Scott Kupor, rulemaking has been held up in the past by the need to summarise tens of thousands of citizen comments. In the case of AI, the agency wants the turnaround to be quicker and more accurate in terms of summaries.
The Department of the Treasury and Department of Commerce are studying AI in detecting tax fraud and patent applications as well as scrutinizing grant submissions. AI has been tested by these agencies in smaller scale environments in the past, but the official use of the GSA will give it the freedom to roll out extensively.Kupor recognized that it is more than the tools to be successful. Oh, we are not getting people who are so familiar with the latest, AI-related things, he said. He continued to say that AI may aid in terms of volume and speed, but experienced human supervision is of importance.
Executive Orders Remake Federal AI Policies and Vendor Requirements
The vendor approvals are only days after President Donald Trump signed three executive orders that are meant to redefine federal AI policy. Among them there is a memo according to which all agencies must buy AI instruments that are not ideologically biased. The order is also consistent with recent conservative outcry against so-called woke AI, or online terminology referring to the notion of progressive bias among commercial AI models.
The GSA affirmed that the enforcement of the executive orders would remain agency by agency where there would be a mandate that every department should interpret and act on the requirement by testing on bias. As political in nature as the administration made their case, the GSA maintained that their vetting was functional and compliance based.Commissioner Gruenbaum said, it is a race. We shall win this race as the president pointed out. His comments reiterated Trump insistence on having America dominate the international push towards AI applications, something which is also becoming evident in federal contracting and planning.
Future and Civilian Vendors as well as Military Signal Spread of AI in Government
Civilian approval of OpenAI, Google, and Anthropic is a turning point but the federal government had been conducting experiments with AI in other sectors beforehand. As an example, the Pentagon has signed prior agreements with OpenAI and Elon Musk xAI in military projects. They are distinct from the GSA role as it helped to illustrate how the government is globally transferring toward incorporating AI into its defense and civil operations.The GSA officials were talking with innuendo that more firms may soon be certified.
All the first three were just further down the review pipeline. “Officials would also consider other cutting-edge AI companies to go with,” officials said.It is more than a regulatory step towards something new but an expected strategic change in the perception of the role of AI in government services in America. It is an indication that there is an increasing support to the notion that with the correct vetting AI has the power to be an effective instrument of governance, fraud detection, innovation, and consultations with citizens.
