Technology & innovation
Mercor, a platform that connects independent contractors with AI companies to train models and chatbots, is facing at least four class-action lawsuits in a California federal court after a March data breach allegedly exposed sensitive personal information of workers on its platform. Meta is among the companies that have paused work with Mercor following the incident.
Plaintiffs allege negligence, breach of implied contract, breach of privacy and violations of California's Unfair Competition Law, claiming Mercor failed to train staff on basic cybersecurity practices.
The breach was reportedly part of a broader attack by a hacking group that exploited a vulnerability in LiteLLM, an open-source interface used by thousands of AI companies.
The number of affected individuals is unclear, but plaintiffs estimate the class exceeds 100 people.
Read more via HR Dive
Senator Bernie Sanders, UAW President Shawn Fain and other labor leaders gathered on Capitol Hill to demand that policymakers take seriously the job displacement risks of artificial intelligence, arguing that workers are being set up for the same broken promises they received when NAFTA passed in the 1990s.
Fain drew a direct parallel between AI and NAFTA, arguing that working people were told free trade would create jobs and that retraining would be available if it didn't. "Instead of all that, we got millions of jobs and thousands of communities destroyed. We got the Rust Belt."
Sanders warned that AI could eliminate jobs across both blue-collar sectors like transportation and manufacturing and white-collar fields including engineering, accounting and law.
AFL-CIO President Liz Shuler urged workers to push for AI protections in union contracts and called on Congress to develop comprehensive safeguards before rapid AI deployment displaces workers across the economy.
Read more via Detroit News
The first lady hosted a global AI education summit at the White House last month, appearing alongside a humanoid robot called Figure 03 to promote AI's role in student instruction. The display drew immediate pushback from teachers' unions, prompting Melania to clarify her position in a follow-up op-ed.
At the summit, the first lady described a future where humanoid AI teachers deliver personalized instruction based on each student's learning speed and emotional state, calling AI the "great equalizer" for students in underserved communities.
Randi Weingarten, president of the American Federation of Teachers, called the robot appearance "every parent's nightmare" at a separate Workers First AI Summit the following day. "This is exactly what Big Tech wants to create: a sense of a society that is being led by and taught by robots," she said.
In a follow-up Fox News op-ed, the first lady sought to walk back the strongest interpretations of her remarks. "AI is not intended to replace teachers with humanoids," she wrote, saying teachers "will remain the foundation of education" while AI supports them with personalized tools and more time for mentorship.
The debate landed the same week New York City released its first AI guidelines for teachers, and as the AFL-CIO convened its inaugural Workers First AI Summit to build a unified labor response to AI's growing impact on workers.
Read more via Reuters, NBC News
The nation's largest school system published a 30-page AI playbook laying out what teachers can and cannot do with the technology across nearly 1,600 public schools. The guide uses a stoplight framework: red for prohibited uses, yellow for proceed with caution, and green for go ahead.
Teachers can use AI for lesson planning, research, drafting communications, scheduling and translation. They cannot use it to assign grades, determine disciplinary action, or craft specialized learning plans for students with disabilities.
The guide reminds teachers that AI should support their work but never replace it, and warns against entering students' private information into AI tools.
The playbook will be updated in June after public feedback. New York City is also exploring the possible creation of an AI-focused high school.
Read more via The New York Times
Meta is reportedly building an AI version of Mark Zuckerberg to advise employees: According to the Financial Times, Meta is training an AI character on Zuckerberg's mannerisms, tone, publicly available statements and views on company strategy, with the goal of having it interact with employees when the CEO can't or doesn't want to. The project follows separate reporting that Zuckerberg is also developing a personal AI agent to help him do his own job. (MSN)
AI-generated Iranian propaganda videos using Lego animations have gone viral: A series of AI-produced animated videos depicting Lego versions of Trump, Netanyahu and other figures spreading anti-American messaging has racked up millions of views across X, TikTok and Instagram after being banned on YouTube. Experts say the campaign, produced by a firm called Explosive Media that acknowledged the Iranian government as a client, represents a new level of sophistication in foreign propaganda. "These new technologies have given Iran an ability to leverage culture in a way that they never have been able to before," said one information warfare expert. (NBC News)
An AI system kept wildlife off Israeli railway tracks with 98% success: A year-long pilot project on Israel Railways used AI-powered alert stations to detect and deter large animals including wild boar, foxes and gazelles from wandering onto tracks. The system, which combines camera-based analytics and automated responses, operates within a 120-meter detection range. Following the successful pilot, Israel Railways signed an agreement to expand the technology across additional sections of its roughly 1,700 kilometers of track. (Stock Titan)
Allbirds is ditching shoes for … AI?: The struggling footwear brand, once valued at $4 billion, announced it is pivoting to AI compute infrastructure under a new name, NewBird AI, after selling its shoe intellectual property last month for $39 million. The company secured $50 million to purchase high-performance chips it plans to lease to customers who can't access AI computing through larger providers. Critics noted the pivot follows a now-familiar pattern of struggling companies latching onto tech buzzwords for a stock bump, pointing to similar moves during the crypto and metaverse hype cycles that ultimately failed. The company's investors are being given a one-year dissolution option if the AI pivot doesn't pan out. (CNBC, Slate)
