In a new study by MIT FutureTech, 272 researchers ranked the top 24 AI risks, write Axios reporters Jim VandeHei and Mike Allen. The researchers “assessed a 12% chance that AI's dangerous capabilities produce a catastrophic outcome by 2030, and another 12% chance of AI-enabled weapons and mass-harm capabilities [with mitigation efforts to reduce risk],” VandeHei and Allen explain.
Researchers from MIT and Toyota Research Institute developed SceneSmith, a system that allows robots to practice everyday tasks and action plans in over 1,300 AI generated environments before entering a real home or workplace, writes Fox News’ Kurt Knutsson. “SceneSmith's ability to combine realistic design with usable physics gives the project its strongest advantage,” notes Knutsson.
Prof. Daniela Rus, director of CSAIL, joins Hakeem Oluseyi, host of GBH’s “Particles of Thought,” to discuss the lab’s work utilizing AI to enhance robots’ ability to move, reason, and act like humans. “The objective is to bring AI’s ability to understand text, images, and other online information to make physical machines intelligent, and if we achieve this then AI will not just reside in our computers—it will drive, it will walk it will fly, it will interact with us in the physical world.”
Prof. Tom Leighton, CEO of Akamai, speaks to Fast Company reporter Victor Dey about his vision for a faster and cheaper AI grid that could serve as an alternative to large data centers, and more effectively route inference workloads. “Many inference tasks do not need the largest model or the largest cluster or the most expensive compute,” says Leighton. “What they need is the ability to marry the right model, in the right place, data and moment, at the most effective cost.”
Graduate student Peter Zhi Xuan Li speaks to Tech Briefs’ Andrew Corselli about his team’s work developing a new chip that allows small, autonomous robots and other battery-limited devices to construct detailed 3D maps of their environments using a fraction of the power required by other systems. “Today, a robot uses separate, incompatible internal representations for different jobs: one for mapping, another for tracking its own motion, another for rendering what it sees, and every conversion adds memory footprint and energy,” says Li. “Our goal is a single shared representation.”
Roughly 100 high school students, two from each state, attended this past weekend’s “Youth AI Festival,” organized in part by MIT RAISE (Responsible AI for Social Empowerment and Education) to discuss the need for clear AI policies for K-12 education, writes Boston Globe reporter Marcela García. “[T]he students discussed, wrote, and approved a model AI-use policy for classrooms, ending in a simulated Senate floor debate and vote at the Edward M. Kennedy Institute,” García explains.
Prof. Daniela Rus, director of CSAIL, joins Hakeem Oluseyi, host of GBH’s “Particles of Thought,” to discuss her group’s work developing liquid neural networks, AI that runs locally on a given device rather than a data center to increase privacy and improve energy efficiency. By using liquid AI, "you avoid privacy concerns and you avoid the security concerns associated with accessing the cloud,” says Rus. “Furthermore, the cost of running your models on a device is much, much lower than running a huge model in the cloud.”
Fortune reporter Preston Fore highlights Principal Research Scientist Andrew McAfee's warning that companies should not replace entry level jobs with AI. “[I]f you’re pulling back on your entry-level hiring, you are probably sacrificing future opportunities to learn and the skilled people of the future,” says McAfee. “You’re also turning off the spigot of the most enthusiastic power users of AI in your organization.”
Prof. Justin Reich speaks to PBS reporter Hannah Grabenstein about the dangers of children using Google’s AI overview, and AI mode. “It's deeply disturbing how poorly these widely accessible tools do,” says Reich. “These are things that technology companies have to fix. And teachers and educators, until this is fixed, probably need to turn a bunch of this off.”
Research Scientist Maria Jesús Saénz discusses the benefits and barriers to companies adopting AI in their supply chains for a Financial Times article by reporter Lucy Colback. “If you are having humans change their process for automation in order to substitute themselves, this is a very perverse thing,” Saénz says. “They want to keep their salary and might boycott the AI. [They may] be algorithm averse.”
Prof. Daron Acemoglu speaks with Washington Post reporter Benjamin Guggenheim about his views on advancing AI, and how the technology might impact the labor market over the next decade. “What we saw at the end of ‘25 and the beginning of ‘26 was an acceleration. I think the agentic AI models are certainly much better in terms of a number of tasks, such as coding and other sort of simple cognitive tasks,” says Acemoglu.
Writing for the Observer, Michael John Gorman, director of the MIT Museum, points to the 1976 Cambridge Experimentation Review Board, a group of citizens tasked with weighing the risks and benefits of DNA splicing technology, as a solution to modern debates over AI. “The dominant assumption in the AI debate is that there are only two options: reckless acceleration or fearful prohibition,” writes Gorman. “Cambridge in 1976 proved there is a third path, and that it runs through the public rather than around it.”
Prof. Alessandro Acquisti speaks with NPR’s Scott Neuman about privacy issues related to surveillance by autonomous vehicle (AV) companies. “There already exist laws that govern duty to report or even duty to protect [for AV companies],” Acquisti says. “The privacy problems arise when and if driverless carrier companies used such laws or ethical obligations as a pretext for blanket, indiscriminate accumulation of identifiable data for unspecified future purposes.”
Prof. Jinhua Zhao takes a car ride through Boston with GBH “Curiosity Desk” host Edgar B. Herwick II to discuss the concerns and benefits of self-driving technology entering the city. “We pay a lot of attention to how good AV [Autonomous Vehicle] safety needs to be, but it’s important for us to understand how bad we have been—the current system is [bad],” says Zhao. “When human drivers are aggressive, they actually indeed pose a challenge on AV—which is that AV needs to learn how to be assertive. If you’re too conservative, you’ll just be stuck there going nowhere.”
CNBC’s Greg Iacurci interviews Prof. Andrew Lo about the potential pitfalls of relying on AI for financial advice. “One of the things about LLMs [Large Language Models] that I find particularly concerning is that no matter what you ask it, it’ll always come back with an answer that sounds authoritative, even if it’s not,” says Lo. “When it comes to very, very specific calculations of your own personal situation, that’s where you have to be very, very careful.”