When AI Gets a New Name, Tech Workers Get New Rules

The Trump administration has ordered federal agencies to stop calling the technology “artificial intelligence” and start calling it “Super Intelligence,” or “SI,” according to reporting by MIT Technology Review. The order, signed at the end of September, has already started showing up in government websites and contract language. For people who write software, sell technology to the government, manage public websites, or prepare bids for federal work, this is not just a vocabulary change. It affects documents, proposals, compliance checklists, and the way workers prove they are following instructions.

An executive order is a written directive from the president to federal agencies about how the executive branch should operate. This one had a single main command: agencies should use the new term instead of the old one. The order says the new wording is meant to show that these systems do more than imitate human thinking. Whether people agree with that reasoning or not, the practical result is that federal workers and contractors now have to decide where the new wording belongs and where older wording can remain.

That matters in government contracting because words in a solicitation can control the work. A solicitation is the government’s official request for companies to offer a product or service, often the first step before a contract is awarded. MIT Technology Review reported that the new term has appeared in solicitations, including an Immigration and Customs Enforcement request for companies to build “SI-powered” surveillance for detention facilities. A vendor responding to that request would need more than technical skill. Its proposal writers, project managers, lawyers, and engineers would all need to use the required language accurately enough to show they understand the customer’s requirements.

This is a real workplace skill. In many technical jobs, especially jobs connected to public money, the correct term is not decoration. It can affect a statement of work, a requirements document, a data policy, a user manual, a training plan, or a cybersecurity review. A statement of work is the document that defines what a contractor must deliver. If the agency says “SI” and a company keeps saying “AI,” the product might be the same, but the bid can look careless or out of step with the customer. In a competitive contract, that can matter.

Inside agencies, the change is uneven. MIT Technology Review found that many government sites have quietly swapped in the new wording, while some pages warn visitors that both old and new terms may appear during a transition. Other parts of government are still using the older language. The Pentagon announced a new warfare study project using the old terminology after the order. A Federal Trade Commission release used a version that did not match the exact two-word, capitalized phrase in the order. Recent Health and Human Services releases showed no sign of the new term, according to the same report.

That uneven rollout creates work. Someone has to update web pages, templates, forms, internal style guides, procurement language, training decks, and archived pages. A style guide is a workplace rulebook for consistent wording, capitalization, and formatting. A content manager might have to decide whether a past document should stay as it was or be revised. A procurement specialist might have to check whether a new bid package uses the new term in every required place. A software team might have to decide whether labels in a user interface should change, even if the code underneath does not.

The change also shows why technical careers often require more than knowing the tool. A machine-learning engineer may build models, which are systems trained on data to make predictions or generate outputs. But in a regulated workplace, that engineer may also sit in meetings with compliance officers, lawyers, product managers, and customers. Compliance means following the rules that apply to an organization, including laws, contracts, and internal policies. If the customer is the federal government, policy changes can quickly become job requirements.

Large technology companies are handling the shift in different ways. MIT Technology Review reported that several major tech leaders signed a nonbinding White House statement on AI safety that used the new term. Nonbinding means it expresses agreement or intention but does not create the same legal duty as a contract or law. Even so, the wording has not fully taken over. Some executives used the new term consistently in earnings calls, which are public calls where company leaders discuss financial results with investors. Others continued to say “AI,” and some used both.

For workers trying to get hired into technology, cybersecurity, digital media, technical writing, or public-sector IT, the lesson is not that one term is automatically better. The lesson is that language can be part of the job. Employers want people who can read requirements closely, adjust to a client’s terminology, document changes clearly, and ask questions when rules conflict. A person who can build a tool but cannot follow the required documentation process may slow a project down. A person who understands both the technology and the paperwork can become valuable fast.

There is also a cost question. MIT Technology Review noted that government employees are spending time writing guidance for departments on how to handle the rebrand, and agency leaders must work on a new legislative definition of “Super Intelligence” for the president by the end of the month. The article does not give a dollar amount for that work. But in any shop, time spent changing labels is time not spent testing systems, fixing bugs, training users, or improving security. That trade-off is exactly the kind of decision workplaces make when a rule changes from the top.