California Governor Gavin Newsom has signed into a law new rules for robotaxis, including one that that's meant to address concerns that they could get in the way of emergency services. Newsom signed Senate Bill 1246, which will establish local penalties for robotaxi companies if their vehicles block first responders, such as ambulances and fire trucks, for more than 30 minutes in case of emergencies. Several events over the past year have heightened people's worries that driverless cars could hinder emergency services from responding in a timely manner. In December 2025, several Waymo vehicles got stuck in the middle of San Francisco roads after a power outage that took out the city's traffic lights. During this year's Fourth of July celebrations, Waymo vehicles also had to be towed away after they ran out of power and got stranded due to heavy San Francisco traffic. Outside of California, in Austin Texas, a Waymo car blocked an ambulance that was responding to a shooting.

The UK's Competition and Markets Authority (CMA) is trying to introduce more search choices where Google has the tightest control: Android smartphones and the Chrome browser. The regulator has proposed allowing users to choose their preferred search provider on those platforms so Google's own search engine isn't used automatically. The CMA also wants to change rules around AI assistants and require search providers to give clearer attribution to publisher content when used. The new proposal follows measures recommended at the beginning of 2026. The CMA's latest proposal has three primary recommendations.

Follow this section to personalize your feed and get instant alerts. Follow Go to your personalized feed WHY FOLLOW? Smart Alerts: Get notified about major news as it happens. Follow this tag to personalize your feed and get instant alerts. Follow Go to your personalized feed WHY FOLLOW?

The score is Machines 1-0 Humans - at least when it comes to predicting football results, anyway. AI - in this case Microsoft chatbot CoPilot - finished top of the BBC predictions league at the first attempt last season, but will it stay there? BBC Sport football expert Chris Sutton, his guests and the BBC readers will be hoping to strike back, and land a blow for the human race. It was a final-day defeat for me, that is how close I came to the title, Sutton said. But I keep reading how AI models are going rogue and attacking each other. So, was I hacked too? It sounds like cheating to me - if so, then last season should really go down as a victory for me.

Driverless cars are becoming more common in some California cities, but when the autonomous vehicles violate traffic laws, police haven't been able to ticket them - until now. The state's Department of Motor Vehicles (DMV) has announced new regulations on autonomous vehicles (AVs), including a process for police to issue a notice of AV noncompliance directly to the car's manufacturer. The new rules, which will go into effect 1 July, are part of a larger 2024 law that imposed deeper regulation on the technology. There have been a number of reports of the cars breaking traffic laws, including during a San Francisco blackout last year. The California DMV is calling the new rules the most comprehensive AV regulations in the nation.

In this paper we study the well-known greedy coordinate descent (GCD) algorithm to solve $\ell_1$-regularized problems and improve GCD by the two popular strategies: Nesterov's acceleration and stochastic optimization. Firstly, we propose a new rule for greedy selection based on an $\ell_1$-norm square approximation which is nontrivial to solve but convex; then an efficient algorithm called ``SOft ThreshOlding PrOjection (SOTOPO)'' is proposed to exactly solve the $\ell_1$-regularized $\ell_1$-norm square approximation problem, which is induced by the new rule. Based on the new rule and the SOTOPO algorithm, the Nesterov's acceleration and stochastic optimization strategies are then successfully applied to the GCD algorithm. The resulted algorithm called accelerated stochastic greedy coordinate descent (ASGCD) has the optimal convergence rate $O(\sqrt{1/\epsilon})$; meanwhile, it reduces the iteration complexity of greedy selection up to a factor of sample size. Both theoretically and empirically, we show that ASGCD has better performance for high-dimensional and dense problems with sparse solution.

That's Harder Than It Sounds Proposed regulations in China would mean the end of flush handles on car doors, with precious little time to roll out the changes. Car door handles seem innocuous. Tesla's electronic, retractable ones--since imitated by plenty of global automakers--have become a symbol of the automaker's willingness to work from design-first principles, reimagining what the car of the future might look like, electric-style. But in September, the National Highway Traffic Safety Administration launched an investigation into the Tesla 2021 Model Y's door handles. More than 140 consumers have complained to the National Highway Traffic Safety Administration (NHTSA) about the door handles, according to a Bloomberg report published last month.

Lifted inference algorithms for probabilistic first-order logic frameworks such as Markov logic networks (MLNs) have received significant attention in recent years. These algorithms use so called lifting rules to identify symmetries in the first-order representation and reduce the inference problem over a large probabilistic model to an inference problem over a much smaller model. In this paper, we present two new lifting rules, which enable fast MAP inference in a large class of MLNs. Our first rule uses the concept of single occurrence equivalence class of logical variables, which we define in the paper. The rule states that the MAP assignment over an MLN can be recovered from a much smaller MLN, in which each logical variable in each single occurrence equivalence class is replaced by a constant (i.e., an object in the domain of the variable). Our second rule states that we can safely remove a subset of formulas from the MLN if all equivalence classes of variables in the remaining MLN are single occurrence and all formulas in the subset are tautology (i.e., evaluate to true) at extremes (i.e., assignments with identical truth value for groundings of a predicate). We prove that our two new rules are sound and demonstrate via a detailed experimental evaluation that our approach is superior in terms of scalability and MAP solution quality to the state of the art approaches.

Retired Charlottesville Fire Chief Charles Werner joins'Fox & Friends First' to discuss drones being used as first responders as Energy Sec. Sean Duffy highlights U.S. drone dominance. Delivery drones could soon take to the skies in full force, following a landmark proposed rule by the Federal Aviation Administration (FAA). The long-anticipated rule is aimed at allowing drones to operate beyond the visual line of sight (BVLOS) -- a move designed to counter China's dominance in unmanned aviation. Currently, operators must obtain individual FAA waivers -- only 657 issued so far -- to fly drones beyond where they can physically see them, hampered by months of delay and bureaucratic setbacks. "Because of that complication, I don't think we saw the innovation that we should have in America," said Transportation Secretary Sean Duffy on Tuesday.

Lifted inference algorithms for probabilistic first-order logic frameworks such as Markov logic networks (MLNs) have received significant attention in recent years. These algorithms use so called lifting rules to identify symmetries in the first-order representation and reduce the inference problem over a large probabilistic model to an inference problem over a much smaller model. In this paper, we present two new lifting rules, which enable fast MAP inference in a large class of MLNs. Our first rule uses the concept of single occurrence equivalence class of logical variables, which we define in the paper. The rule states that the MAP assignment over an MLN can be recovered from a much smaller MLN, in which each logical variable in each single occurrence equivalence class is replaced by a constant (i.e., an object in the domain of the variable).