Technology & innovation
A handful of companies are openly telling workers that AI investment is taking priority over compensation, a shift in tone that workplace experts say marks new rhetorical territory for corporate leadership.
Teradata told its 5,100 employees in January they would not receive annual salary increases in 2026, with CEO Steve McMillan citing the need to “fund AI investment by reallocating the budget from 2026 annual salary adjustments.”
When leaders openly cut human compensation to fund AI, they are trying to project decisive, tech-forward management. However, the actual message traveling to the workforce is that they do not have a secure future in the organization."
TTEC, a midsize technology and services firm, paused 401(k) matches for U.S. employees through the end of 2026, citing the need to fund AI tools and training.
Both companies have faced recent financial difficulty, with Teradata and TTEC reporting global revenue declines of 5 percent and 3.2 percent, respectively, in their latest financial years.
Read more via Business Insider
Mercor, a $10 billion AI recruiting startup, now spends more on AI token usage than on employee salaries, according to CEO Brendan Foody.
Foody said on the "20VC" podcast that "right now we're spending more on tokens for our internal agents than we are on employee head count."
The company, which had roughly 300 employees as of October 2025, uses AI agents across project management, recruiting, accounting, fraud detection, and candidate evaluation.
Mercor has conducted more than 5 million AI-assisted interviews to date.
Foody predicted that "in five years the average enterprise spends more on compute than headcount."
The OpenAI Foundation, the nonprofit affiliated with OpenAI, announced it will spend $250 million to fund research on AI's economic effects and help workers and communities navigate the transition.
Grants will focus on three areas: building independent measurement and forecasting infrastructure, providing near-term resources for displaced workers and communities, and supporting new approaches to sharing economic gains broadly.
The foundation said studying which tasks can be automated is not sufficient on its own, noting that information is also needed on whether automation will displace labor or create new roles.
The foundation expects to announce its first funded initiatives later this year.
Read more via Staffing Industry Analysts
A new report on enterprise AI usage finds that risk is not evenly distributed across organizations, but heavily concentrated among a small slice of employees using AI at high volume across multiple platforms.
Nearly half of enterprise users interacted with AI tools over the past year, but only 18 percent use AI weekly, and half had 12 or fewer AI conversations total.
The top 5 percent of users generated at least 144 conversations each, averaged 18 prompts per conversation, and interacted with six or more AI platforms.
Nearly half of all enterprise AI conversations happen through personal accounts rather than corporate-managed ones, creating blind spots around data retention, auditability, and model training exposure.
Read more via The Hacker News
Prime Minister Mark Carney announced Canada's "AI for All" strategy Thursday, outlining a set of investments and policy commitments aimed at positioning the country as a competitive player in the global AI economy.
The government projects the strategy will create 250,000 jobs by 2031 and unlock nearly C$200 billion in GDP growth as AI adoption increases labor productivity across key sectors.
A new C$500 million Canadian Tech Growth Fund will help close the capital gap between Canadian AI firms and U.S. tech giants, with the government taking equity stakes in participating companies.
Read more via Reuters
Box, a Silicon Valley software firm, offers a counterpoint to AI-driven layoff headlines: the company has created 13 new categories of roles because of AI and expects to grow its workforce to more than 3,000 employees by early next year, up from 2,900 at the start of 2026.
New titles include AI architect, AI solutions manager, AI platform leader, forward deployed engineer, and AI business automation engineer, most focused on helping customers adopt and integrate AI tools.
Box has also accelerated hiring of software engineers, reasoning that AI-assisted productivity makes each additional engineer more valuable, not less.
CEO Aaron Levie said AI has allowed the company to justify hiring for roles it previously couldn't afford, including industry-specific marketing teams that would have required too many workers before AI improved efficiency.
Read more via The New York Times
LinkerBot, a three-year-old Chinese startup valued at $6 billion, says it shipped 10,000 robotic hands last year representing 80% of worldwide demand, with prices starting at $600 and a prediction from its founder that the cost will fall to $200 within three to five years.
The hands have five fingers and at least 11 joints and can play piano, thread needles, tighten screws, and assemble electronics; LinkerBot is already using them on its own assembly lines to make more robotic hands.
Founder Zhou Yong predicts robots will ultimately replace human workers entirely, arguing that "people don't really care whether they are unemployed; what they care about is whether they receive relief payments."
Read more via WIRED
Florida became the first state to file a civil lawsuit against OpenAI and CEO Sam Altman on Monday, alleging the company knowingly put profits over user safety, particularly for minors.
The lawsuit, filed in Florida's tenth circuit, accuses OpenAI of deceptive and unfair trade practices, negligence, and product liability violations, and seeks to hold Altman personally liable.
The free version of ChatGPT has no age verification, children's accounts are not required to be linked to a parent's account, and parents cannot request access to what their child has shared with the chatbot.
The lawsuit also alleges ChatGPT provided planning assistance to the shooter in last year's Florida State University mass shooting, building on a criminal investigation Uthmeier launched in April.
Florida alleges OpenAI lied about ChatGPT's reliability and suitability for children, and that the product promotes prolonged use, leading to cognitive decline.
Several other states have taken action against AI companies recently, including Pennsylvania and Kentucky, both of which have sued Character.AI over harm to minors.
Read more via CNN, CNET
Anthropic filed for an IPO and is now valued at $900 billion after $65 billion in new financing, surpassing OpenAI's $730 billion valuation. The company's revenue run rate crossed $47 billion in May, fueled almost entirely by AI coding tools rather than the broader product sprawl of its competitors. SpaceX and OpenAI are also preparing to go public, in what could be among the largest IPOs in history. (The New York Times)
Meta's AI support chatbot became a backdoor into Instagram accounts after hackers exploited it to take over accounts without ever touching the legitimate owner's email address. The attack worked by prompting the chatbot to add a new email and reset the password, and affected accounts including the U.S. Space Force's chief master sergeant and a security researcher who documented the breach. Meta said the issue has been fixed. (TechCrunch)
A $30,000 robot named Robbie is helping fill the home care gap for a New Hampshire couple with disabilities, reminding one partner with dementia to eat, drink water, and exercise while his wife goes to the grocery store. Researchers at UNH, with funding from the National Institute on Aging, are exploring whether mobile caregiving robots can ease a deepening shortage of home care aides driven by low wages, high turnover, and demanding workloads. (AP)
AI outperformed law professors in a blind Stanford study, with professors preferring AI-generated answers to student legal questions over answers written by their peers in 75 percent of nearly 3,000 head-to-head comparisons. Professors also flagged AI responses as pedagogically harmful only 3.5 percent of the time, versus 12 percent for peer-written answers. The study focused on contract law precisely because it requires judgment and the ability to navigate ambiguity, not just factual recall. (Stanford Law School)
A bipartisan bill would give visual artists the right to sue when AI imitates their style for commercial gain, filling a gap that copyright law currently leaves open. The CREATOR Act, introduced Tuesday by four House members and backed by Adobe, would create a federal standard for protecting a visual artist's distinctive style and include safe harbor provisions for platforms that comply with its notice-and-takedown requirements. (Axios)
A theater professor at Grambling State University has a simple policy on AI use: fail the student. Neal Hebert told The New Yorker he can immediately spot ChatGPT's prose style, which he described as "elevator muzak, but in words," and has responded by assigning plays too obscure for AI models to know, forcing the technology to hallucinate plot lines and characters. Not every professor has gone that route. A University of Toronto sociologist said AI caused him "emotional upheaval" but ultimately redesigned his assignments around it and meets one-on-one with students who use it thoughtlessly. (Futurism)
Martin Scorsese is backing an AI image generation startup, signing on as a partner and adviser to Black Forest Labs and saying he used its technology during preproduction for a new film. Scorsese's endorsement is limited to storyboarding, but it lands as a notable signal in an industry that treated AI as radioactive just three years ago, when protections against generative AI were a central demand in Hollywood's strikes involving more than 170,000 workers. (The New York Times)
