DataBreachToday.com

DeepSeek-R1 Model Rivals OpenAI at Fraction of Cost, Challenges US AI Leadership
The Chinese DeepSeek-R1 model rivals top U.S. AI models in performance while dramatically lowering training costs. Built on lower-end Nvidia GPUs, R1's efficiency reshapes the AI landscape, bolstering AMD's prospects and reshaping AI infrastructure discussions in Silicon Valley and beyond.

Experts Explain Complexity of Company's Massive Data Breach Analysis Endeavor
UnitedHealth Group says its previously eye-popping estimate of 100 million people affected by last year's attack on its Change Healthcare unit nearly doubled in recent months to a staggering 190 million victims. Why did it take so long for the company to figure out the extent of the compromise?

Europe Targets Officers of Unit 29155 of the Russian Main Intelligence Directorate
The European Union sanctioned on Monday three officers of a Russian military intelligence unit for their role in cyberattacks targeting Estonia in 2020. The sanctions are tied to the unit's 2020 hack of the Estonian government departments and the exfiltration of sensitive documents.

China's DeepSeek AI Model Sparks US Policy Debate Amid Growing Industry Concerns
The global artificial intelligence race saw a historic "Sputnik moment" this week when the Chinese startup DeepSeek claimed to develop a competitive model with $6 million and a stockpile of old Nvidia semiconductors - defying export restrictions and raising alarms about China's ability to innovate.

Series C Funding Targets Supply Chain Risks in AI, Next-Gen Infrastructure Security
Eclypsium raised $45 million in Series C funding to address emerging cyber threats in AI workloads, GPU systems and the global supply chain. With support from Qualcomm and 1011 Ventures, the company will to tackle advanced cyberthreats from nation-state adversaries like Volt Typhoon.

AI Startup's R1 Model Draws Praise and Skepticism
An open reasoning model from Chinese artificial intelligence startup DeepSeek has the tech industry assessing its potential impact as shares of U.S. technology mainstays plummeted in trading on Monday. Hangzhou-based DeepSeek released its R1 model on Jan. 20.

Preauthentication Deserialization Flaw Could Result in Remote Code Execution
Software vendors and national security agencies are urging immediate patching of a critical SonicWall flaw days after the security device manufacturer disclosed that hackers are actively exploiting a zero-day. The flaw doesn't require user authentication.

RansomHub Theft Hit Patients of 2 Dozen HCF Facilities and Home Healthcare Unit
A chain of more than two dozen skilled nursing and rehabilitation facilities is notifying tens of thousands of patients whose information was compromised in a hacking incident last fall. Russian-speaking cybercriminal gang RansomHub claims to have published 250GB of data stolen in the heist.

Undersea Cable Damage in Baltic Sparks Concerns of Russian 'Shadow Fleet' Activity
Swedish authorities seized a vessel in the Baltic Sea as part of a criminal investigation into sabotage of an undersea cable between Sweden and Latvia after a series of regional cable disruptions, including one involving suspected Russian shadow fleet activity, raising security concerns.

Why MFA and Data Minimization Remain Key for Preventing Massive Data Breaches
While PowerSchool's investigation into the massive theft of its customers' data is continuing, clear lessons have already emerged. Count among them the importance of using multifactor authentication, which could have safeguarded access to PowerSchool's exploited customer support systems.

With the pace of global change so often creating a sense of accelerating chaos, it's easy to view cyber defenders as firefighters constantly on call. But Black Hat conference founder and creator Jeff Moss warned that "things have been on fire for as long as I can remember."

While artificial intelligence platforms and tools promise to offer encouraging potential in healthcare, many are unprepared to deal with the risks these emerging technologies pose - similar to the early days of social media, said Keith Fricke, partner and principal of tw-Security.

Chinese Hackers Hitting Unpatched Products From Microsoft, Sophos, Fortinet, Ivanti
Chinese nation-state hackers who surreptitiously gained "broad and full" access to telecommunications networks in the U.S. and dozens of other countries have regularly exploited known flaws in their networking gear that the victims failed to patch, security experts have warned.

Operator Cannot Yet Reliably Perform Complex, Customized Tasks
OpenAI introduced an AI agent capable of independent action with the launch of Operator, an general-purpose AI tool that interacts with websites to perform tasks. The agent can navigate menus and complete forms to do tasks such as travel booking, ordering takeout, buying stuff or scheduling tasks.

Fraud Expert Ken Westbrook on Successful Ways to Stop Fake Investment Sites
According to the FBI, losses from investment scams surged 38% between 2022 and 2023. Fraudsters are using highly effective tactics, including sending text messages to lure victims to fake cryptocurrency platforms, said Ken Westbrook, founder and CEO of Stop Scams Alliance.

The EU-US Data Privacy Framework Requires a Functional PCLOB
A Trump administration move to gut a key oversight body meant to guarantee European data rights in the United States could endanger the legal basis underpinning commercial data flows across the Atlantic. The board is charged with overseeing U.S. surveillance practices.

Under the Trump administration, the proposed update to the HIPAA Security Rule - issued in the final weeks of the Biden administration - is likely to get trimmed but not totally cut, predicts regulatory attorney Sharon Klein of the law firm Blank Rome. What else should the health sector expect?