The US OMB releases new AI guidance, requiring that all federal agencies submit an annual AI report and have a senior leader overseeing all AI systems they use
Given the widespread usage of machine learning in software this sounds so broad as to be useless. … X: Vice President Kamala Harris / @vp : At the first-ever Global AI Summit last year, I laid out our vision for a future where AI advances the public interest. To help build that future, I am announcing our first government-wide policy to promote the safe, secure, and responsible use of AI. https://www.whitehouse.gov/... Steven Sinofsky / @stevesi : FACT SHEET: Vice President Announces OMB Policy to Advance Governance, Innovation, Risk Management in Federal Agencies' Use of AI // WH hates “Big Tech” so much that put in place policies that in many cases will definitely do opposite of stated goals. https://www.whitehouse.gov/... Suresh Venkatasubramanian / @geomblog : The set of domains considered rights-impacting covers all the areas where we've seen AI used and misused and is extremely comprehensive. 4/n [image] Suresh Venkatasubramanian / @geomblog : The set of domains considered safety-impacting is equally impressive - and health care appears in both! 5/n [image] @amprog : Today, @VP announced new Office of Management and Budget (OMB) guidelines for federal agencies on the use of artificial intelligence (AI). @Alondra Nelson, distinguished senior fellow at the Center for American Progress, led the drafting, public and interagency review, and— @datasociety : Today @OMBPress released its memo on the US govt's use of AI. Across its various mechanisms, its core principle is this: When the risks to people's rights outweigh the benefits, AI should be off the table. 1/3 https://www.whitehouse.gov/... @cendemtech : Today, @OMBPress has released the final version of its guidance on federal agencies' use of #AI. @CenDemTech supports the Biden Administration's efforts to promote more effective & efficient government use of AI. https://cdt.org/... @civilrightsorg : “Today, the OMB's guidance takes us one step further down the path of facing a technology rich future that begins to address its harms.” Click here to read our full statement on the new OMB memo on agency implementation of the AI executive order. https://civilrights.org/... John E. Mitchell II / @johnemitchellii : We've long said that the U.S. must lead the world in AI innovation. The new WH/OMB's policy of establishing Chief AI Officers to oversee AI adoption & use across government operations is an important step in helping the U.S. become a global leader in AI development. Nik Marda / @nrmarda : Today, the U.S. government is establishing new rules for its own use of AI, and they're a HUGE deal Buckle up folks — we're going to unpack how they work and why they're so good 🧵 https://blog.mozilla.org/... Jalen Brown / @itsjalenb : NEW: The White House is bolstering artificial intelligence safeguards to mitigate privacy and discrimination risks in healthcare. The guidance mandates human oversight of AI-generated diagnostic decisions and establishes AI governance boards across agencies—and other measures. @brennancenter : New rules for AI use by federal agencies include important safeguards but “give agencies too much leeway to opt out,” says Faiza Patel. “The stakes are too high for the government to merely pay lip service to protecting civil liberties and rights.” https://www.brennancenter.org/ ... Nik Marda / @nrmarda : Federal agencies will now: ✅ Pursue high impact AI uses ✅ Require bias to be mitigated for many uses of facial recognition ✅ Protect federal workers from AI surveillance in the workplace ✅ Incentivize companies who sell AI to the government to also follow these rules Kevin Bankston / @kevinbankston : I'm hopeful the requirements on agencies here will succeed in forcing government investment—of money and brains—into developing a heretofore seriously lacking level of practical specificity around responsible AI norms and best practices. So much of that work has typically... Suresh Venkatasubramanian / @geomblog : We've moved from asking WHETHER we should deploy responsibly, to asking HOW to deploy responsibly. The AI Bill of Rights spelt the HOW out in great detail, and the OMB memo now codifies this for the entire US government. 11/n, n=11. Bill Nelson / @senbillnelson : Artificial Intelligence (AI) is one of the most powerful tools of our time. Today, the @WhiteHouse announced its plan to safely integrate AI across the government. At NASA, we'll continue using AI responsibly to study our climate, revolutionize air travel & explore the universe. Suresh Venkatasubramanian / @geomblog : An Executive Order gets the attention, the fancy signing ceremonies, and the coverage. The OMB memo is where the rubber meets the road. When I was asked about the EO last year, I kept saying, “it's great, but let's wait for the OMB memo to come out”. 2/n https://www.whitehouse.gov/... Pamela Melroy / @astro_pam : Today, the @WhiteHouse issued its first-ever government-wide policy to mitigate the risks & harness the benefits of AI. At NASA, we'll continue to leverage AI safely for the betterment of humanity—including our missions to study climate change and explore the Moon, Mars & beyond. @ombpress : Today, OMB issued #AI policy requiring federal agencies to adopt concrete safeguards to protect people's rights and safety, while also seizing the promise of AI for public good. https://www.whitehouse.gov/... Barack Obama / @barackobama : This is important work being led by President Biden and Vice President Harris. We can drive innovation while also creating the right safeguards and oversight of artificial intelligence. Suresh Venkatasubramanian / @geomblog : Today is a very important day for AI governance in the US. The @OMBPress released their instructions for all Federal agencies on how to protect our rights when considering the use of AI, following up on Oct 2023 EO on safe, secure, and trustworthy AI. https://www.whitehouse.gov/... 1/n LinkedIn: Lili Gangas : Today the The White House announced OMB policy that took into account public comment on AI Executive Order and provided an update on all tasks completed to date … Saeed Al Dhaheri : The White House issued a government-wide policy to mitigate the risks of using AI in federal operations. Some important highlights: … Gareth Stokes : For those in any doubt that getting on top of AI governance issues should be a top priority for any business, interesting to see this mandate for US federal agencies to appoint chief AI officers... … Augie Ray : Does your organization have #AI policies? If not, you're late. And if so, who are they most designed to protect—your company or your customers? …
Context & Ripple Effects
This turns earlier White House efforts to address AI risks, including plans for agency safeguards and new research, into operating requirements for federal agencies. It also moves beyond the voluntary commitments AI companies made to the White House in 2023.
The significance is administrative rather than technical: agencies must identify and account for the AI systems already embedded in their work, with a named senior owner responsible for oversight.
First-order effects
- Federal agencies must produce annual accounts of their AI use and assign a senior leader to oversee their AI systems, creating a clear internal point of accountability.
- Agency AI deployments now face a common reporting and governance process rather than being handled solely within individual program teams.
Second-order effects
- Agencies will need inventories, ownership records, and review workflows for AI systems in order to meet reporting and oversight obligations; this can expose fragmented or poorly documented deployments.
- Technology suppliers serving agencies may face greater demand for documentation and governance support as customers need to explain and oversee the systems they use.
Third-order effects
- Federal AI policy is shifting from broad safety principles toward operational AI governance: recurring reporting, named accountability, and controls built into deployment processes.
- If applied consistently, this model can make governability a more important criterion in public-sector AI adoption, though its practical force will depend on how agencies implement and enforce the guidance.
The trend: AI governance is moving from voluntary commitments and high-level principles toward accountable, repeatable operating controls inside major institutions.