|
|
|
|
|
ILLUSTRATION: DADO RUVIC/REUTERS
|
|
|
|
|
Elon Musk's xAI lost a bid to pause California's AI law mandating that companies disclose a summary of the data used to train their AI products. xAI challenged the law, which went into effect Jan. 1, in Los Angeles federal court. The company said the law forces it to reveal trade secrets and violates free-speech rights. (Reuters)
|
|
|
|
Anthropic’s Claude sniffs out a bevvy of bugs. It took Anthropic’s most advanced AI model about 20 minutes to find its first Firefox browser bug during an internal test of its hacking prowess. Over two weeks in January, Claude Opus 4.6 found more high-severity bugs in Firefox than the rest of the world typically reports in two months, said Mozilla, Firefox's parent organization. (WSJ)
|
|
|
Cyber incident at the FBI: The Federal Bureau of Investigation confirmed it is investigating “suspicious activities” in an internal, unclassified system related to surveillance and investigations, the agency said in a notice to Congress. The FBI first spotted the activity, which exploited the agency's network security controls and an external internet provider, Feb. 17. (Associated Press)
|
|
|
The Internal Revenue Service is reviewing allegations that it improperly shared private taxpayer. The agency recently was accused of giving data beyond the scope of its agreements to Immigration and Customs Enforcement, and other potential privacy missteps. (FedScoop)
-
"A thorough review of all cyber is undergoing right now, and we are also … having outside reviews," IRS head Frank Bisignano told lawmakers.
|
|
|
Local government hacked: Phone and some tech systems remained down Thursday in Passaic County, N.J. government offices after a cyberattack on Tuesday. (NJ.com)
|
|
|
|
|
90
|
|
Number of zero-day vulnerabilities Google tracked last year, up from 78 in 2024. Forty-three of 2025's total affected enterprise software, up from 36 in 2024.
China-linked hacking groups remain the biggest exploiters of zero days, Google said.
|
|
|
|
|
|
|
|
|
|
PHOTO: YVES HERMAN/REUTERS
|
|
|
|
|
The European Union should brace for heightened threats of terrorism and cyberattacks on critical infrastructure related to the conflict in Iran, Europol warned. Polarizing rhetoric online about the war could stir radical actors in and outside of the EU, the police agency said. (Reuters)
|
|
|
|
X threatens to dock the pay of influencers who post AI-generated material about armed conflict without labeling it as such. Offenders will be suspended from X’s creator revenue-sharing program for three months, said Nikita Bier, head of product for the social network. (TechCrunch)
|
|
|
🎧 New episode: Oxford Analytica's Laura James updates us on the conflict in Iran and outlines scenarios if fighting continues. Also, the head of New York's financial watchdog outlines her priorities. Perry Cleveland-Peck hosts.
|
|
|
|
|
|
|
|
60
|
|
Number of pro-Iran hacktivists active since Israel and the U.S. bombed the country Feb. 28, according to threat researchers at Palo Alto Networks. As of March 2, that total includes pro-Russia groups loosely coordinating with Iran-linked hackers.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
