U.S. President Donald Trump signed an executive order on Wednesday making it easier to fire 8,000 of some of the best-paid government workers, part of a broader effort to overhaul the federal workforce.
The order, released by the White House and the Office of Personnel Management, strips job protections from a mostly senior group of federal workers earning up to almost $200,000 a year and who are deemed to be "influencing" government policy.
In a call previewing the move, Scott Kupor, director of the Office of Personnel Management, which oversees the government's human resources policies, said the administration needs to employ people willing and able to carry out orders to achieve the administration's policy priorities.
"You can have any political views, but if you allow those views to basically interfere with your willingness to actually carry out lawful orders and policy directives with the administration, then this provides a mechanism obviously for people in those agencies to be able to be removed effectively at will," he said.
The order shows Trump is persisting in his efforts to discipline and fire career employees who he sees as undermining his political goals, a year after billionaire Elon Musk left his post overseeing an effort to slash government spending and payrolls.
Trump believes his agenda was hampered by career federal workers who opposed his policies during his first term.
The number of workers affected by the order is well below a ceiling estimate of up to 50,000 workers who could have been made subject to the new rules. Senior administration officials on the call said Trump could expand the grouping, but has no immediate plans to do so.
Federal worker unions and their allies sued in January to stop the policy before it was fully developed. Federal judges paused the litigation while the Trump administration finalized changes.
Bank of America (BAC.N) said on Wednesday it would welcome about 4,000 summer interns and full-time campus hires, signaling a commitment to entry-level talent even as the rise of AI fuels industry concerns over the future of junior banking roles.
The bank will make 2,000 full-time recruits and open 2,000 internship roles, according to an executive at BofA.
Comments from senior banking executives underscore growing industry anxiety over the disruptive potential of AI, which is increasingly automating complex, data-intensive tasks previously handled by human staff.
Wall Street's integration of generative AI has cast a shadow over entry-level banking roles. However, BofA said it was committed to additional entry-level hiring initiatives.
Uber recently disclosed that it burned through its entire 2026 AI budget in the first four months of the year. 🆘 🔥
At the time, 95% of engineers were using AI tools monthly, and roughly 70% of committed code was AI-assisted.
That’s not an AI failure story. It’s an ROI story, and one that most don’t know how to measure well.
For the last two years, the conversation has been about how quickly and efficiently AI can write code.
Now the question is more important: Is all that extra code creating real value, or just more vibe-coded slop? More output does not always mean more impact.
And as token usage climbs, companies will need a much clearer way to distinguish real leverage from expensive, often company-wide coding activity.
Who is measuring AI development effectiveness well? Or at least has a solid grip on how to think about ROI in this new coding era, especially as it starts to look more expensive than the traditional human-centric way of development and coding?
At the time, 95% of engineers were using AI tools monthly, and roughly 70% of committed code was AI-assisted.
That’s not an AI failure story. It’s an ROI story, and one that most don’t know how to measure well.
For the last two years, the conversation has been about how quickly and efficiently AI can write code.
Now the question is more important: Is all that extra code creating real value, or just more vibe-coded slop? More output does not always mean more impact.
And as token usage climbs, companies will need a much clearer way to distinguish real leverage from expensive, often company-wide coding activity.
Who is measuring AI development effectiveness well? Or at least has a solid grip on how to think about ROI in this new coding era, especially as it starts to look more expensive than the traditional human-centric way of development and coding?