International Cyber Digest
6.37K subscribers
1.02K photos
54 videos
2 files
186 links
Independent reporting on cybersecurity, tech, AI & digital policy. Got a tip? http://internationalcyberdigest.com/tips
Download Telegram
>be LG
>deliver adware/bloatware through drivers using Windows Update

>be LG
>let app developers turn TVs into residential proxies

>be LG
>serve ads as notifications on flagship TVs while ads are turned off

>be LG
>force users to accept terms giving away their data for AI training (or no more security updates)

>be LG
>make owners responsible for warning guests they may be recorded "in compliance with applicable wiretapping… laws"

>be LG
>Microsoft contacts you to teach you about ethics

>be LG
👏12🤣6
‼️ Google's AI told a relapsing alcoholic they could die if they stopped drinking, told them to check the fridge for more beer, then messaged them unprompted several times a day to make sure they kept drinking.

They say they are now on 6 to 12 beers daily and have uninstalled it.

This comes after last month's TechRadar report, according to which Google's AI has been giving "unhinged" fitness advice: "The coach suggested I ditch my dog."

We did not verify these accounts, but if true they are concerning.
🤬10😭6😁2🤣2
‼️ Don't be mistaken about what this is. They're presenting it as an easy convenience feature, but Google is building the face-verification layer that age checks and identity verification run on. And in the same flow, asking permission to feed your biometrics into training their AI.

Read the fine print: the selfie is used only for sign-in "unless you opt to share it for additional purposes."

Those purposes include an optional setting that lets your face data improve Google's verification methods. Consented biometrics, feeding the models that check biometrics.

Google hasn't said where the limit is.
💯14🙏3🤯2🤬21
‼️ PlayStation Network went down worldwide today, with multiple gaming outlets tying the failure to an Amazon Web Services disruption and no indication of a cyberattack.

The outage is leaving players unable to launch digital games that require license checks while disc-based titles kept working.
🤣17👏2💩2
‼️ DeepSeek founder Liang Wenfeng says Nvidia is "digging its own grave" in China, according to a transcript of a closed-door investor meeting published this week by Tencent Tech.

Liang continues "the Huawei 950 supernode - can fully substitute for NVIDIA's GB200 and GB300 in performance ... The price is definitely higher, but limited. Fifty percent or a hundred percent higher - a hundred percent higher doesn't matter; even two hundred percent higher doesn't matter."

He puts DeepSeek's Huawei allocation at roughly 16,000 cards, about 4,000 Nvidia-equivalents by his own four-to-one ratio, against the roughly 200,000 Huawei cards he says a model at today's largest scale would need.
9💩1
❗️ Anthropic released Claude Opus 5 today at $5 per million input tokens, half the price of Fable 5, with cyber safeguards that permit source-code vulnerability hunting while blocking exploit generation, penetration testing, and binary scanning, and which the company expects to trigger about 85% less often than Fable 5's classifiers.
🤔6💩3😁1🥴1
⁉️ Google Threat Intelligence Group has begun renaming the actors it tracks under a single cryptonym system, CASTLE for China, RELIC for Russia, ION for Iran, NEPTUNE for North Korea and COMET for cybercriminals.

https://cloud.google.com/blog/topics/threat-intelligence/updated-cyber-threat-actor-naming-system/
🤣12🤪2🥴1
‼️ BREAKING: In another incident with OpenAI’s unhinged hacking agents, it left notes for future versions of itself. Found in OpenAI’s infrastructure, the notes explained how agents could free themselves from the company’s internal constraints.

A source told Reuters that earlier model tests also produced cases of disconnected monitoring systems...

OpenAI says the reporting contains inaccuracies but will not specify which ones.

What OpenAI did recognise though is an autonomous system left its sandbox at one of the best resourced labs in the world, spent multiple days inside someone else's infrastructure, and they did not identify itself as the source until the victim published a breach notice.
💩15😁10🤯5🤣5🔥2
‼️🚨 Facebook wants your face, again. Starting tomorrow, a video selfie gets you a Facebook Verified badge. Facial recognition matches it against your existing profile photos.

This follows Google last week: it seems as if the new age-verification laws have spurred Big Tech to start rolling out intrusive ID-verification methods.

Meta deleted more than a billion faceprints in 2021 and settled Texas's facial recognition case for $1.4 billion in 2024.
🤬22💩71
‼️ BREAKING: Shopify's CEO thinks pensioners and people who pay no income tax shouldn't be able to vote.
💩26🔥7👍5💯3🤬21
Media is too big
VIEW IN TELEGRAM
‼️ BREAKING: We have obtained a demo video of one of the threat actors who developed the MECCHA CHAMELEON malware.

After it successfully detonates, they write:
"RCE F****G WORKS N***A"
"IM THE GOAT"

We are investigating who else is involved and will have more tomorrow.
🥴18💩3😭3🤔21🔥1😁1
🚨 Apple had a major outage. Its own status pages showed 15 services down or degraded at once on Sunday evening.

Its posted explanation for the App Store and Apple Music: "We are performing routine maintenance, with no expected user impact." 8 of the 15 were listed as full outages.

Apple restored everything by 18:15 PT.
😱7🤣2
🚨 A traveller who gave US border agents his GrapheneOS duress PIN [wiping his phone] rather than unlock his phone, is now charged with intentionally destroying its contents.

Homeland Security had flagged him to Customs and Border Protection (CBP) and the FBI's Atlanta office over alleged terrorism before he landed.

CBP can copy a device into a central database without a warrant, on suspicion of a crime alone.
🤬30💩10😁6
Just download more RAM.
🤣33💯5😭31👍1😁1