This morning, Google announced a $50 million commitment to train more than 300,000 skilled trade workers across the US including electricians, fiber technicians, the people who build and maintain data centers. It follows Meta's $115 million America's Workforce Academy announced on Monday, complete with paid training and a job guarantee for every graduate. Anthropic has a $150 million fellowship. Add BlackRock, Lowe's Companies, Inc., and Ford Motor Company, and corporate America has now pledged over half a billion dollars to the skilled trades in just a matter of months.

So what's actually going on with these investments?

These companies have figured out that they cannot win the AI displacement argument. "Is AI taking jobs?" is an unwinnable question, because the honest answer is yes, some, and on purpose. No amount of messaging fixes that. So instead of defending against a question they can't win, they're funding a question they can which is: who's investing in American workers?

This new workforce story is for Washington, for the public, and for the communities where data centers need permits, power, and political goodwill. Look at where Meta's pilot programs are: Louisiana, Ohio, Indiana, Texas. Those aren't random; it's where the data centers are, and they're geographies where "Big Tech is training your electricians" buys something no amount of lobbying can.

Importantly, this works because it's real. These aren't pledges to a foundation that eventually evaporate into the ether. Meta's program pays people while they train and guarantees them a job. Google's money goes to 14 labor unions and four trade associations. The operational truth is what makes the comms strategy hold weight.

Every company making AI-driven cuts right now is making an implicit claim about its future: we're becoming something different, and this is what it takes. A lot of companies make that claim with language alone ("transformation," "the AI era," "positioning for what's next") and then wonder why employees and journalists don't believe them. Meta and Google are making the same claim with money, programs, and commitments that can be checked. The argument about their future is backed by what they're building, not by how they describe it.

If your company is making AI-related cuts and the rationale is the future of the business, the question, as always, is about what you can point to. What are you funding, building, or guaranteeing that proves the future you're describing is real? If the answer is nothing, or if the entire case for the transformation lives in copy, then no version of the message will hold up, because there's nothing under it.

Asha Sharma became XBOX CEO in February, after Phil Spencer retired from Microsoft . On Wednesday, Bloomberg News' Jason Schreier reported that Xbox was planning major layoffs shortly after the fiscal year closes at the end of June, citing people familiar with the company's strategy, along with significant cuts to marketing and other parts of the business.

The picture Bloomberg painted wasn't flattering. The division was described as "over-extended," and the financial reality behind that characterization was stark. Excluding the Activision Blizzard King acquisition, Xbox spent over $20 billion on content, platform, and hardware subsidies over five years while annual revenue declined by nearly half a billion dollars in that same period.

But rather than go quiet and let the story run, shortly after the Bloomberg piece went live, Sharma and chief content officer Matt Booty published an Xbox Wire post titled "Next 100 Days: Xbox Reset", the full text of which had just been emailed to Xbox staff globally. The memo didn't dodge the numbers - it confirmed the 3 percent accountability margin, the $20 billion in investment against declining revenue, and the blunt conclusion that, "going forward, this cannot continue."

Bloomberg's story ran anyway BUT it ran alongside Sharma's own account rather than without it and that's a really meaningful distinction. Because once Bloomberg had the gist of the story, her options narrowed to two which was 1) say nothing and let Bloomberg's sources define the narrative, or 2) put her own words into the public record simultaneously and own the financial diagnosis herself. She chose the latter and almost every outlet that followed ran her direct language. "Over-extended." "Hard truths." "Reset." Those are her words, not a reporter's characterization. The memo, published publicly rather than left to leak, gave the story a protagonist with a point of view instead of a subject being written about.

Most executives in that situation go quiet and hope the story is smaller than they fear. But Sharma did the opposite. That's not a communications trick but a decision about whether you trust your own voice more than you fear what happens when you use it.