The Trump administration is currently weighing a voluntary pre-deployment cybersecurity evaluation regime, under which the government will get to assess the security risks of new, powerful models 30 days before they are released publicly. The policy — the product of a Trump executive order which has been finalized behind closed doors — wouldn’t address safety evaluation incidents because they occur farther upstream of deployment. “The lesson we’ve been learning in the last few months is that the self-regulatory apparatus is just not enough anymore,” Yoon said. “There are competitive pressures
The AI safety test is becoming a safety risk | TechCrunch… Zoox told TechCrunch in a statement that the software update “enhances the existing capability of detecting active emergency scenes by adding the ability to detect and respond to heavy smoke in certain situations.” Nobody was on board the vehicle during the June incident, and Zoox told the National… …
… The proposed legislation, expected to be introduced in Congress this week, follows a number of incidents in which autonomous vehicles have blocked ambulances and fire trucks, driven into active crime scenes, and failed to recognize safety signals such as traffic cones and flares. …
… NHTSA will gather additional information and data, and solicit public comments, to determine if a rulemaking for a new safety standard is appropriate under the requirements of the Vehicle Safety Act.” The decision comes less than a year after NHTSA opened an investigation into Tesla’s door handles … …
… The policy — the product of a Trump executive order which has been finalized behind closed doors — wouldn’t address safety evaluation incidents because they occur farther upstream of deployment. “The lesson we’ve been learning in the last few months is that the self-regulatory apparatus is just not… …
… On the podcast, the OpenAI co-founder called it an “extremely sci-fi cyber incident… t his is the first security incident that I have felt very viscerally.” OpenAI researchers have paused training on that model while they work out how to keep their sandbox secure. …
… Campbell pointed to an incident where Microsoft deployed a version of the company’s GPT-4 model in India through its Bing search engine before the model had been evaluated by the company’s Deployment Safety Board DSB . …
… If it may never be possible to know with certainty that a model is fully aligned, then the practical question comes down to how to safely contain and control increasingly capable systems. “There’s not yet a good understanding of how to align the most capable AI systems, but there’s much more consen… …
… All of the crashes have come while the Avride vehicles were under the supervision of a safety monitor in the driver’s seat. Reached for comment, Avride declined to explain why the safety monitors did not intervene in these crashes. …
… Federal Aviation Administration spokesperson Steve Kulm said the agency has not received a report about the incident. The FBI said it was aware of the incident and is working with local and corporate partners on the matter. …
… The company claims that every time it receives this kind of notification, the incident is reviewed by a human. “We strive to review these safety notifications in under one hour,” the company says. …