The New Oil 1 year ago

#Google says its new #AI models can identify emotions — and that has experts worried

TechCrunch

Google says its new AI models can identify emotions — and that has experts worried | TechCrunch

Google is releasing a new open model family, PaliGemma 2. Experts are concerned about its emotion recognition capabilities.

#privacy

The New Oil 1 year ago

#CopilotVision, #Microsoft’s #AI tool that can read your screen, launches in preview

TechCrunch

Copilot Vision, Microsoft's AI tool that can read your screen, launches in preview | TechCrunch

Microsoft's newest AI tool, Copilot Vision, can read your screen — or rather, the sites you're browsing with Edge. It's now in pre...

#privacy #Windows #Copilot #cybersecurity

The New Oil 1 year ago

A New Phone Scanner That Detects #Spyware Has Already Found 7 #Pegasus Infections

WIRED

A New Phone Scanner That Detects Spyware Has Already Found 7 Pegasus Infections

The mobile device security firm iVerify has been offering a tool since May that makes spyware scanning accessible to anyone—and it’s already tu...

#iVerify #malware #cybersecurity

The New Oil 1 year ago

New website shows you how much #Google #AI can learn from your photos

Ars Technica

New website shows you how much Google AI can learn from your photos

Upload your photo and get a thorough, three-paragraph description of it.

#privacy #Ente

The New Oil 1 year ago

#Amazon and #Google Must Keep Their Promises on #ProjectNimbus

Electronic Frontier Foundation

Amazon and Google Must Keep Their Promises on Project Nimbus

When a company makes a promise, the public should be able to rely on it. Today, nearly every person in the U.S. is a customer of either Amazon or G...

#privacy #AI #politics

The New Oil 1 year ago

#Ente: #MachineLearning is live

ente

Machine learning is live

Ente's privacy preserving ML is live now

#cybersecurity #privacy #AI

The New Oil 1 year ago

#RaceCondition Attacks against LLMs

Schneier on Security

Race Condition Attacks against LLMs - Schneier on Security

These are two attacks against the system components surrounding LLMs: We propose that LLM Flowbreaking, following jailbreaking and prompt injection...

#LLM #AI #cybersecurity

The New Oil 1 year ago

#Tuta User Survey 2024: Your guidance for Tuta's future development

Tuta

Tuta User Survey 2024: Your guidance for Tuta

We asked the Tuta community for their insights so we can better plan the 2025 roadmap. Check out the results!

#privacy #cybersecurity #email #AI #PostQuantumEncryption

The New Oil 1 year ago

Someone Made a Dataset of One Million #Bluesky Posts for 'Machine Learning Research'

404 Media

Someone Made a Dataset of One Million Bluesky Posts for 'Machine Learning Research'

A Hugging Face employee made a huge dataset of Bluesky posts, and it’s already very popular.

#AI #privacy

The New Oil 1 year ago

#SUSE unveils major rebranding, and a new #AI platform that protects your data

ZDNET

SUSE unveils major rebranding, and a new AI platform that protects your data

SUSE has been busy! The European Linux power wants you to know it's also a major cloud and open-source player in North America and beyond.

#FOSS #privacy #Linux

The New Oil 1 year ago

#UK seeks collaboration for security research lab to counter #Russia and ‘new #AI arms race’

TechCrunch

UK seeks collaboration for security research lab to counter Russia and 'new AI arms race'  | TechCrunch

The U.K. is seeking collaboration for a new AI security research lab, designed to counter the growing threat of hostile states.

#cybersecurity #politics

The New Oil 1 year ago

#PlayAI clones voices on command

TechCrunch

PlayAI clones voices on command | TechCrunch

PlayAI, a platform to generate voice clones and voiceovers, has raised capital in a large seed funding round.

#privacy #AI #deepfake

The New Oil 1 year ago

#Microsoft is finally testing its #Recall photographic memory search feature. It’s not perfect

CNBC

Microsoft is finally testing its Recall photographic memory search feature. It's not perfect

Microsoft on Friday finally started letting people test Recall, a so-called photographic memory search feature for the latest Copilot+ PCs. It's no...

#privacy #cybersecurity #AI

The New Oil 1 year ago

It's Surprisingly Easy to #Jailbreak #LLM-Driven Robots

IEEE Spectrum

It's Surprisingly Easy to Jailbreak LLM-Driven Robots

Researchers induced bots to ignore their safeguards without exception

#AI #cybersecurity

The New Oil 1 year ago

#Microsoft rolls out #Recall to #Windows Insiders with #Copilot+ PCs

BleepingComputer

Microsoft rolls out Recall to Windows Insiders with Copilot+ PCs

​Microsoft announced today that its controversial AI-powered Recall feature is finally rolling out to Windows Insiders in the Dev Channel using S...

#WindowsRecall #AI #privacy #cybersecurity

The New Oil 1 year ago

#PaloAltoNetworks warns cybercriminals are breaking into its customers’ firewalls — again

TechCrunch

Palo Alto Networks warns hackers are breaking into its customers' firewalls — again | TechCrunch

Hackers have compromised potentially thousands of Palo Alto customers by exploiting two new zero-day vulnerabilities

#cybersecurity

The New Oil 1 year ago

#TMobile actually stopped some cybercriminals from stealing customer data

The Verge

T-Mobile actually stopped some hackers from stealing customer data

Didn’t have that on my bingo card.

#privacy #cybersecurity #DataBreach

The New Oil 1 year ago

#Microsoft accused of creating a monopoly on #US government systems through free upgrades

TechRadar

Microsoft accused of creating a monopoly on US government systems through free upgrades

Did Microsoft lock in the US government?

#cybersecurity #politics

The New Oil 1 year ago

#UK MPs ignore concerned constituents questions about #FacialRecognition Technology

Privacy International | Action Privacy International

#privacy #politics

The New Oil 1 year ago

#Fitness app #Strava is tightening third-party access to user data

Ars Technica

Fitness app Strava is tightening third-party access to user data

Training apps, AI use, and trend analysis seem to be off-limits in Strava's API.

#privacy #AI