Microsoft trying to stop Copilot generating fake Putin comments on Navalny's death
Plus: US Department of Justice has hired its first chief AI officer
AI in brief Microsoft is investigating fictional press statements about the death of political prisoner Alexei Navalny, written by its AI Copilot and falsely attributed to Russian president Vladimir Putin.
Lawyer and de facto opposition leader Navalny died on February 16th while serving a 19-year sentence for extremism-related charges, on top of another 11-and-a-half year sentence for fraud. His death sparked protests in Russia, and events honoring his life around the world.
When a journalist at Sherwood Media, the publishing arm of financial services company Robin Hood, prompted Microsoft's Copilot to generate text for a news article about Navalny's death, the service fabricated statements. It claimed that US president Joe Biden held Putin responsible for Nalvalny's death, and that, in response, Putin called the accusations "baseless and politically motivated."
Putin has not made a public statement about Navalny’s death.
"We have investigated this report and are making changes to refine the quality of our responses," a Microsoft spokesperson told Sherwood Media. "As we continue to improve the experience, we encourage people to use their best judgment when viewing results, including verifying source materials and checking web links to learn more."
Regulators halt Waymo's self-driving car expansion plan
The California Public Utilities Commission has temporarily halted Waymo expanding its self-driving car service beyond San Francisco.
The Google spinout wanted to operate its driverless cars down Highway 101 to reach the Peninsula all the way down through Millbrae, San Mateo, Palo Alto and Sunnyvale in the Bay Area. Waymo also wanted to cover a large patch of Los Angeles – from Santa Monica to Monterey Park, down from Manhattan Beach to Paramount.
But those plans have been put on hold until June at least, as regulators continue mulling over the expansion. San Mateo county supervisor David Canepa said the decision to hold Waymo back was made after the startup failed to provide enough information on safety risks.
"Since Waymo has stalled any meaningful discussions on its expansion plans into Silicon Valley, the CPUC has put the brakes on its application to test robotaxi service virtually unfettered both in San Mateo and Los Angeles counties," Canepa declared in a statement, according to the San Francisco Standard.
"This will provide the opportunity to fully engage the autonomous vehicle maker on our very real public safety concerns that have caused all kinds of dangerous situations for firefighters and police in neighboring San Francisco."
Earlier this month, a Waymo car hit a cyclist at a four-way intersection. Thankfully, the accident wasn't too serious, and the cyclist walked away without any major injuries.
- Simon Willison interview: AI software still needs the human touch
- Did all that AI chatbot hype boost Bing's market share? Oh, wait, never mind
- Cruise being investigated over car crash that dragged victim along the road
Computer scientist hired as DoJ's chief AI officer
The US Department of Justice has hired its first-ever chief AI officer to source technical expertise and advice as the technology increasingly impacts law enforcement.
Jonathan Mayer, an assistant professor at Princeton University's Department of Computer Science and School of Public and International Affairs, will take on the newly created role. He will be responsible for guiding attorneys in complex cases involving technical subjects – such as AI, cyber security, and more.
"The Justice Department must keep pace with rapidly evolving scientific and technological developments in order to fulfill our mission to uphold the rule of law, keep our country safe, and protect civil rights," attorney general Merrick Garland explained in a statement this week.
"Jonathan's expertise will be invaluable in ensuring that the entire Justice Department – including our law enforcement components, litigating components, grantmaking entities, and US Attorneys' Offices – is prepared for both the challenges and opportunities that new technologies present."
Mayer will also lead the DoJ's Emerging Technology Board and liaise with other governmental agencies on AI and law. He has an academic background in computer science, policy, national security, and consumer protection. ®