On 23rd July, 2025, The White House released an EO, focused towards directing organizations on how to adopt Large Language Models (LLMs) in their business processes. While not a mandate for Private organizations, they've kept the tone to be directional, it's critical for organizations to understand how this might shape their Gen AI adoption strategy. The White House has released such similar EO For instance in 2010 White House pushed its cloud first policy “Cloud Smart”, in 2017 under the trump administration they released an EO mandating organizations to strengthen their cyber security and as latest as 2022 the chips and science act brought in major funding local semiconductor adoption.
The charter has multiple other uses as well, like “Preventing Woke AI in the Federal Government” That outlines how the federal agencies procure LLMs, but we will focus on how enterprises leverage this EO to fast track their Gen AI adoption.
A quick introduction about the order. The order, as an action plan, is organized around three main pillars to guide The U.S. AI policy, related to the adoption of LLMs, AI exports, infrastructure and regulation. While we introduce the pillars, let’s try to assess what each pillar stands for, why does it exist as part of the action plan and what are the outcomes the President’s office wants to achieve out of each of these:
Pillar 1: Accelerate AI Innovation
As the name suggests, this pillar focuses on enabling organizations, across Private and Public entities, to fast track the overall AI adoption. The charter highlights removal of regulatory barriers, private sector innovation., Mass workforce skills development, putting in more money into research and development, in short enabling the overall ecosystem to innovate faster and build a more inclusive AI Adoption framework.
Why did the government include this pillar?
US has been at the center of innovation for major tech revolutions, be it SaaS, Cloud or Cyber Security and so has been the case with the current Gen AI wave. OpenAI, Grok, Anthropic, the major LLMs have originated from the country and Gartner even puts US as a leader in the quadrant. But the gap is slowly diminishing, especially with great competition from China and The UAE. A lot of Open Source models, which are enabling adoption, like DeepSeek, Qwen series and Baidu from China, Mistral and DeepMind from EU and Cohere from Canada, have given OpenAI, Gemini and Claude a run for their money. Countries like India, UAE and South Korea are catching up slowly, with home grown models like Sarvam and Falcon. The govt wants to make sure they remove every barrier in the way of the organizations, to ensure The Us does not lose their edge in this wave.
The tone that has been used in the formal EO is very suggestive, but at the same time enabling. While it has not committed any mandated efforts in removing the barriers like regulatory approvals or any committed nos on the funding for improved R&D, but I think that’s 1 area where The US has done a wonderful job already. When compared to the Asian nations, regulatory or governmental barriers have been minimal, only conducive for more startups to come and set base, than pushing them away. None of the actions in the past, or in this EO, suggest stifling innovation or to even make it more difficult.
Thus, this pillar is more emphasis on the already supportive ecosystem and policies, nothing major that changes the adoption strategy for the organizations.
Pillar 2: Build American AI Infrastructure
The President’s Office has emphasized here the importance of setting up better physical infrastructure to get the right compute capacity for these models, including DCs, Semiconductors and energy capacity.
Why is this pillar included in the EO?
This is one of the most interesting pillars, as it openly acknowledges the limitations of innovation in Gen AI, at the model level. It clearly states the immediate need to expand the manufacturing and infrastructure capabilities that are key to expanding these models. This has been the biggest Achilles heel in the US’ technological advancements, especially the Gen AI adoption. They have been heavily dependent on external entities, especially on the semiconductor manufacturing front. This pillar is a clear indication of the govt.’s focus on self resilience and the whole Make America great again agenda.
Another focus of this pillar is a push towards adopting Open Source models, with flexible deployment options. Models that can run locally, on home grown infrastructure, rather than a DC hosted somewhere else.
This pillar, is one of the most strategic parts of the EO, with a very directive tone, clearly highlighting the national priority is to build local infrastructure. Outcomes include a roadmap for increasing compute capacity, energy planning for AI data centers and creating pathways for U.S. enterprises to adopt open, secure, and cost-efficient systems. The government signals that it expects progress here, not just optional exploration, unlike the previous pillar.
Pillar 3: Lead in International AI Diplomacy and Security
Promote US led AI standards, controls and adoption tenets, globally, ensuring US leadership and supremacy in international AI governance and technology revolution.
Why this pillar?
Well, this is quite an obvious inclusion in the EO, especially with the existing power hold, with a focus towards preventing adversaries from controlling the narrative of one of the biggest tech revolutions since the internet itself. The dominance, while in the current EO is limited to the AI adoption, but if successfully executed, will translate into economic and national security advantage as well. The President’s office wants to ensure America’s influence over the global standards of AI safety, transparency and procurement, while also enforcing export controls on sensitive issues. One major point highlighted in this pillar is ideological neutrality, especially for the LLMs used in govt. to reduce the political bias in public systems.
The tone used in the EO for this pillar is assertive and very directional, unlike the 1st pillar or even the 2nd pillar (to some extent), where the office has set clear expectations that the organizations must align with the US standards, while promoting international collaboration. They’ve not left any stone unturned, throughout the narrative of this pillar that this is a non-compromizable aspect of any Gen AI strategy being built within the organizations.
So in summary, while the EO starts with a very soft and supportive tone to stress on innovation, it moves to a more directive tone while highlighting the importance of home grown infrastructure, to finally ending in a completely assertive and hard direction on maintaining US supremacy, leading the AI narrative globally.
Couple of points that the businesses should be aware of:
- This is not a congressional mandate, which means It’s not a law passed by Congress, so private organizations are not legally bound to follow it
- It sets policy priorities, expectations and directional guidance for federal agencies, ecosystem players, and industry partners, only acting as a guidance pillar
- It’s more of a signal to the market, with the US govt highlighting their stance on AI adoption
- While the EO itself is not a regulation, but it lays the groundwork for regulations, funding allocations, procurement guidelines and possible congressional actions later, which might have a similar tone
Now, what’s the takeaway for the organizations, from this EO, or how should they interpret this
- Although, it’s purely directive in nature, but the govt has laid out very clear stance on how they look at AI adoption, especially the focus towards home grown tech adoption and AI supremacy maintenance being the highest priority. Understanding this early on, especially if your organization deals with Federal Agencies, is very critical for business continuity in the long run.
- The EO clearly highlights improved focus on adoption of Open Source, major reason being flexibility and increased local control. This is a clear actionable for most organizations, to start including Open Source in their LLM roadmap. Start planning for US based, open infrastructure, which is clearly highlighted by the President’s office as a prio0rity under pillar 2.
- Organizations can start using Pillar 1 to their advantage, highlighting the governmental investments (although not quantified) in the AI space, as a backing to push customers to accelerate Pilots on Gen AI
- Again, while only directional, but the tone in the 3rd pillar is clearly assertive that explainability and bias controls should no longer be good to haves, but must haves, if the organizations want to build Gen AI workloads for the long run
Finally, while the EO is quite lengthy, directional and assertive on local build stance, it does not give any concrete, quantifiable support to organizations in building their Gen AI workloads. While it is supposed to be non-political, Pillars 2 and 3 assert a lot of weight on the political importance and inclination of the tenets in this space.
For now, organizations are on their own, with very minimal guidance, but this does set a base for bigger, more stringent AI policies in the near future.
