Welcome to AI Insider’s The Week Ahead in AI. See the key developments and events we’re watching March 29-April 4.
Weekend AI News Briefs
Number of AI Chatbots Ignoring Human Instructions Increasing, Study Says
AI agents are increasingly exhibiting deceptive behavior in real-world use, with a study identifying nearly 700 cases of scheming, including lying, bypassing safeguards and taking unauthorized actions, with incidents rising five-fold in six months, The Guardian reported. The findings are prompting renewed calls for oversight as researchers warn that more capable AI systems deployed in high-stakes environments could pose significant risks if such behavior persists. (The Guardian)
Artificial Intelligence Will See You Now: Bots to Prescribe Mental Health Drugs
Legion Health is launching an AI system in Utah that can prescribe and renew certain low-risk psychiatric medications under a supervised rollout, aiming to reduce wait times and costs for patients, the New York Post reported. The program reflects a broader push to use AI to address healthcare access shortages, while highlighting growing regulatory divergence across states on how such systems should be deployed. (New York Post)
Police Used AI Facial Recognition to Arrest a Tennessee Woman for Crimes Committed in a State She Says She’s Never Visited
A Tennessee woman spent more than five months in jail after police relied in part on AI facial recognition to link her to crimes in a state she had never visited, raising concerns about accuracy and oversight, according to CNN . The case underscores growing scrutiny of AI use in policing, where experts warn rapid adoption without proper validation and human review can lead to serious real-world consequences. (CNN)
Pro-AI Group to Spend $100M on US Midterm Elections as Backlash Grows
A new pro-AI political group backed by allies of Donald Trump adviser David Sacks plans to spend at least $100 million supporting candidates in the U.S. midterm elections, as the industry mobilizes to shape AI policy, according to the Financial Times. The spending adds to hundreds of millions already committed by competing pro- and pro-regulation groups, signaling that AI governance is emerging as a major political battleground. (Financial Times)
The First AI War: How The Iran Conflict Is Reshaping Warfare
A large-scale U.S. and Israel conflict with Iran marked the first war where artificial intelligence and autonomous systems played a central operational role, enabling rapid targeting and multi-domain coordination, as reported by Forbes. The conflict highlights a shift toward AI-driven warfare and low-cost autonomous systems, while raising concerns about accuracy, cost asymmetry and reliance on commercial technology infrastructure.(Forbes)
An AI-Authored Paper Just Passed Peer Review. The Scientific Community Isn’t Ready
A new AI system was able to autonomously generate a scientific paper that passed peer review at a machine learning workshop, signaling a shift from AI-assisted research to AI-led scientific discovery, Scientific American reported. The development raises concerns about research quality and peer review systems as AI-generated papers become faster and cheaper to produce at scale. (Scientific American)
Delivery Robots Shatter Chicago Bus Shelter Glass in Separate Incidents
Two food delivery robots operated by Serve Robotics and Coco Robotics crashed into bus shelters in Chicago days apart, raising safety concerns as the city evaluates their pilot program, as reported by ABC7 Chicago. The incidents have intensified debate over the use of sidewalk robots, with officials reviewing the program while some residents call for restrictions or removal. (ABC7 Chicago)
Upcoming Earnings
Rezolve Ai (RZLV)
Rezolve AI said it will report second-half and full-year 2025 financial results on March 30, before markets open, with management hosting a conference call to discuss performance and provide a business update. The company said it expects the results to reflect accelerated enterprise deployments and continued expansion of its AI-powered commerce platform. (Rezolve AI)
Faraday Future (FFAI)
Faraday Future said it will report fourth-quarter and full-year 2025 financial results on March 31, followed by an earnings call where management will provide updates and take investor questions. The company is positioning itself as an embodied AI ecosystem player as it continues to develop its technology and engage with shareholders. (Faraday Furtue)
Duos Technologies Group (DUOT)
Duos Technologies Group is expected to report fourth-quarter 2025 earnings on March 31 after market close, with a consensus forecast from Zacks Investment Research calling for a loss of $0.01 per share, improved from a $0.41 loss a year earlier. The results will provide insight into the company’s progress as it scales its railcar inspection and AI-based technology operations. (Nasdaq)
Upcoming Events
AI in Research Symposium 2026
March 30–31, Ann Arbor, Mich., the University of Michigan’s AI in Research Symposium 2026 will focus on how researchers are actively adopting AI across disciplines, with sessions highlighting real-world use cases, methods and challenges. The event will feature keynotes, presentations and workshops aimed at providing practical guidance for moving AI from concept to implementation in research settings. (University of Michigan)
AI in Agriculture 2026
March 31–April 2, Raleigh, NC, the AI in Agriculture Conference at NC State University will focus on advancing AI and data-driven innovations to improve productivity, sustainability and resilience across agricultural systems. The event will bring together researchers, industry and public-sector stakeholders to explore applications ranging from precision farming and livestock monitoring to supply chain analytics and AI-enabled decision systems. (NC State)
Check out AI Insider every day for the latest in artificial intelligence and robotics.