Cyber criminals are still attempting to exploit the coronavirus pandemic for their own gain and they’re being helped by website templates that allow them to mimic government agencies and companies. Researchers at cybersecurity company Proofpoint have identified over 300 phishing campaigns designed to steal personal information and bank details from victims – and many are using sites that are indistinguishable from the real thing, complete with authentic imagery and user interfaces. The security company warned that these template make it easy for scammers to quickly create high-quality, malicious web domains to insert into their COVID-19 phishing campaigns.
In a paper published this week on the preprint server Arxiv.org, scientists at DeepMind introduce the idea of simple sensor intentions (SSIs), a way to reduce the knowledge needed to define rewards — functions describing how AI ought to behave — in reinforcement learning systems. They claim that SSIs can help to solve a range of complex robotic tasks, for example grasping, lifting, and placing a ball into a cup with only raw sensor data. Training AI in the robotics domain typically requires a human expert and prior information.
Federal law enforcement officials said Monday they had unlocked the iPhones of the perpetrator of a December terrorist attack at a U.S. Naval base — and sharply criticizing Apple for not granting them access to those encrypted communications. FBI technicians cracked the phones of a Saudi aviation student who killed three U.S. sailors at the Naval Air Station Pensacola, uncovering evidence linking him to an Al Qaeda affiliate, Attorney General William Barr said. Barr and FBI Director Christopher Wray urged Silicon Valley companies to write software that allows investigators to access encrypted communications with a warrant, a move that technology firms and security experts have criticized for years.
In a recent technical paper, researchers affiliated with the University of Southern California and Amazon Robotics explored a solution to the problem of lifelong multi-agent path finding (MAPF), where a team of agents must be moved to constantly-changing goal locations without collisions. They say that in experiments, it produces “high-quality” solutions for up to 1,000 agents, significantly outperforming existing methods. MAPF is a core part of a number of autonomous systems, like driverless vehicles, drone swarms, and even video game character AI. No doubt of interest to Amazon is its applicability to warehouse robots — as of December, Amazon had more than 200,0000 mobile machines inside its fulfillment network.
Microsoft and FedEx announced a joint, multiyear partnership on Monday that the pair believe could help “transform commerce” through FedEx’s logistics network and Microsoft’s cloud. The two said their first service, called FedEx Surround, will give real-time analytics into supply chain and delivery, so companies could potentially better ship goods. The companies did not disclose the full nature of the partnership, including how much FedEx will be paying Microsoft to use its Azure cloud technology. But the partnership represents yet another example of Amazon rivals choosing to go with Microsoft’s cloud offering over the cloud computing market leader Amazon Web Services (AWS).