Millions of working Americans struggle to read at a functional level — and artificial intelligence may be helping hide it.

Why it matters: Low literacy is quietly becoming a major economic drag, even as AI tools allow workers to complete tasks they may not fully understand.

• Experts warn that this can mask deeper skill gaps until workers are asked to make judgments, solve problems or evaluate AI-generated answers.
• Some researchers call this "cognitive surrender" — when people defer to AI outputs without fully evaluating them.
• That creates a workforce that looks productive on the surface but is vulnerable to disruption.

By the numbers: Roughly 130 million U.S. adults read below a sixth-grade level, according to adult literacy estimates.

• About 43 million U.S. adults cannot read, write or do basic math above a third-grade level, according to ProLiteracy.
• More than 90% of jobs require some form of computer literacy, Sharon Bonney, CEO of the Coalition on Adult Basic Education, told Axios.

Zoom in: Low literacy at work is showing up in emails, safety instructions, training materials, math-heavy trades, health benefits forms and computer-based tasks.

• Bonney said adult education programs often see learners who want better jobs but lack the basic reading, math, English-language or digital skills needed to enter apprenticeships, community college or higher-paying work.
• "If you can't read, write, speak the language, can't use a computer, your chances of being gainfully employed are pretty slim," Bonney said.

What they're saying: "The net effect of AI on the workplace is probably going to be increased demand and need for workers with higher levels of basic skills, not lower," Stephen Reder, professor emeritus of applied linguistics at Portland State University, tells Axios.

• "If it's flashing a red warning light that says we have a literacy challenge, then we probably really do have a literacy challenge," Amanda Bergson-Shilcock, senior fellow at the National Skills Coalition, told Axios.

Yes, but: Americans have not stopped buying books.

• Independent bookstores have grown in recent years, and Barnes & Noble has staged a comeback, suggesting reading culture remains strong for some.
• However, Reder said book buying and literacy skills are not the same thing. The bigger divide may be between people who use reading deeply in everyday life and those who rarely practice those skills.
• "Not only are skill levels going down," Reder said, "but particularly among people at the lower end of the skill spectrum, the amount that they use the skills that they have is going way down."

Behind the scenes: Workers have long found ways to hide literacy gaps, like asking family for help, avoiding written tasks or relying on coworkers, Bergson-Shilcock said.

• Now, AI may be accelerating that — creating what she calls an "invisible drag on productivity" that doesn't show up in data but slows teams down.
• In some cases, Bergson-Shilcock said low literacy among supervisors can ripple across entire workplaces, affecting performance and compliance.

The bottom line: AI may help workers keep up, but it also raises the risk that they're producing answers they don't fully understand.

• Reder compared AI to calculators, which made math easier but did not eliminate the need to understand what problem you were solving.
• "You still need to know what you're doing," he said.

As AI changes the way the world gathers information, some critics say that it is perpetuating stereotypes and erasing cultural nuances for Indigenous groups and people of color.

Why it matters: Most mainstream models are trained on the work of Western writers — particularly white men — and regularly mimic those values, writing styles, viewpoints and biases.

• Some critics say the data grab is a new form of colonialism, where information gathering replaces Imperial-era land seizures while the AI companies — rather than a conquering nation — reap profits from marginalized groups.
• Data collection from these groups is often done without their consent or any verification that the information is accurate.

What they're saying: "Colonialism is always portrayed as something that happened in the past ... many countries got independence, and then the textbooks say 'colonialism is over,'" Julian Posada, a Yale assistant professor who studies the relationship between human labor and data production, tells Axios.

• Posada says that modern-day colonialism still exists, but people often fail to recognize it.

Context: Most large language models are made by the WEIRD — Western, Educated, Industrialized, Rich and Democratic societies — and pull data from social media, websites, news archives, and digitized materials that largely originate in North America and Europe.

• Those training materials have resulted in LLMs inventing details based on Western assumptions about cultural traditions or values, and those errors persist despite Big Tech putting in work to train them with more diverse viewpoints and data.

Case in point: Aditya Vashistha, an assistant professor at Cornell University, tells Axios that AI models will often say all Indian food is "rich and aromatic and spicy" even though some isn't, flattening the diversity of the Indian palate.

• "You will find different regional cuisines which differ in the spices which are used, or in what moderation, like the amounts they use."

Zoom out: Taking the data itself is a "deeply colonial act," Nick Couldry, co-author of "Data Grab: The New Colonialism of Big Tech and How to Fight Back," tells Axios.

• "To say, 'Well, it's just out there. We can just take it.' That was what colonialism was about, just taking everything."

Keep reading.

President Trump is pursuing a Bernie-like interest in having the U.S. government take stakes in AI giants — not out of populism, but with a dealmaker's eye for profit.

Why it matters: Having the government take equity in private companies — anathema to free-market capitalism and Republican dogma about "picking winners and losers" — has become a defining feature of Trump's second term.

• The U.S. government now owns shares of chipmakers, miners and quantum computing companies, often taking a stake in exchange for federal money that was originally meant to come with no ownership strings attached.
• Trump said Friday that he "should be a stockbroker," pointing to his administration's recent deal to take a stake in Intel.

Keep reading.

• OpenAI plans to revamp ChatGPT to make it more appealing to enterprise customers, according to sources. (Financial Times)
• Sriram Krishnan plans to leave his role as a White House AI adviser at the end of the month. (TechCrunch)
• Mark Gurman has an illuminating look at a key Apple meeting that led to the Siri update expected to debut later today. (Bloomberg)

Ina again, wishing a very happy 80th birthday to my mom. It was lovely to be able to celebrate her this weekend at Salt & Stone, a great restaurant if you are ever in Sonoma County.