⚠️ This week’s Threatsday Bulletin is here.

🔹 CarPlay exploit (CVE-2025-24132)
🔹 Root access—no clicks needed
🔹 Patch released, but OEMs haven’t applied it
🔹 and more critical threats you can’t ignore...

Your systems may already be at risk.

Read ↓ https://thehackernews.com/2025/10/threatsday-bulletin-carplay-exploit.html

🚨Pentest findings are still being buried in PDFs. By the time they reach IT, attackers may already be in.

The twist? The fix isn’t more reports—it’s automation. Real-time ticketing, instant alerts, zero lag.

The 7 workflows every security team should automate ↓ https://thehackernews.com/2025/10/automating-pentest-delivery-7-key.html

🚨 Malicious PyPI package spotted: “soopsocks” hit 2,653 downloads before takedown.

Advertised as a SOCKS5 proxy, it secretly:
– Dropped a Go-based backdoor (_AUTORUN.EXE)
– Ran PowerShell + VBScript
– Changed firewall rules
– Exfiltrated system data

Full report ↓ https://thehackernews.com/2025/10/alert-malicious-pypi-package-soopsocks.html

🚨 A single click on a shortcut file was enough.

Confucius hackers just stole passwords, screenshots & files across Pakistan’s government and defense networks.

They hid the breach behind routine Windows .LNK files — a trick most security tools missed.

Read ↓ https://thehackernews.com/2025/10/confucius-hackers-hit-pakistan-with-new.html

🚨 U.K. just ordered Apple to break iCloud encryption.

The target? Every Briton’s backups.

Officials want far more than just disabling Advanced Data Protection—secret orders demand access to entire categories of iCloud data.

How deep does this go? ↓ https://thehackernews.com/2025/10/threatsday-bulletin-carplay-exploit.html#u-k-demands-access-targets-britons-backups

🚨 Ordinary PDFs just became hacker weapons.

A new “MatrixPDF” toolkit turns any file into a phishing lure. Even legit docs now hide fake secure prompts + JS redirects.

One click = stolen creds or malware payload.

Details ↓ https://thehackernews.com/2025/10/threatsday-bulletin-carplay-exploit.html#normal-pdfs-turn-into-malware-traps

🚨 Google Drive now pauses itself when it detects ransomware.

The AI spots mass file encryption attempts—then halts syncing before your data gets nuked.

It learned its tricks from millions of real victim files.

How it works → https://thehackernews.com/2025/10/threatsday-bulletin-carplay-exploit.html#drive-pauses-syncs-buys-you-minutes

#threatsday #cybersecurity

🚨 CISA just flagged a live zero-day in Smartbedded Meteobridge.

Remote attackers can hijack weather stations → full root access.

No login needed—just click a link. CGI script + eval = instant code execution.

Patches only dropped in May ↓ https://thehackernews.com/2025/10/cisa-flags-meteobridge-cve-2025-4008.html

🚨 A new APT group is hijacking real government email accounts to hack Russian state agencies.

Their malware exfiltrates data through a Telegram bot.

And it’s not just Russia—English & Arabic filenames suggest the target list is much wider.

Read ↓ https://thehackernews.com/2025/10/new-cavalry-werewolf-attack-hits.html

⚡ Enterprises are hitting a new blind spot.

Passwork 7 now combines password + secrets management in one platform.

That means SSH keys, tokens, and DB logins live next to everyday credentials — all gated by roles and vault design.

One misstep = organization-wide exposure.

Here’s how it works ↓ https://thehackernews.com/2025/10/product-walkthrough-how-passwork-7.html

🕵️ One click. One ZIP.

SORVEPOTEL self-replicates through WhatsApp Web — blasting itself to all your contacts and groups until your account is banned.

Targets enterprises, uses LNK → PowerShell for persistence.

Read more → https://thehackernews.com/2025/10/researchers-warn-of-self-spreading.html

⚠️ Rhadamanthys now fingerprints devices and hides malware inside PNG/JPEG/WAV files.

It’s sold by subscription from $299/mo — a business, not a hobby.

Expert detail: payloads need a secret key from the C2 server to decrypt (stego delivery).

Read the stealth trick that beats many scanners → https://thehackernews.com/2025/10/rhadamanthys-stealer-evolves-adds.html

💀 Another breach caused by a “secure” password.
Hackers don’t need zero-days when your policy is the backdoor.

See why complexity rules fail—and how to block breached creds before attackers use them.

📅 Join the live webinar. Save your spot → https://thehacker.news/password-graveyard

🚨 Detour Dog just flipped the script!

The group once known for shady redirects is now pushing Strela Stealer via hacked WordPress sites + DNS TXT records.

90% of sites look normal—until they quietly fetch malware.

Learn more ↓ https://thehackernews.com/2025/10/detour-dog-caught-running-dns-powered.html

🧨 500% surge in scans hitting Palo Alto Networks logins.

GreyNoise says it’s the highest spike in months — and eerily mirrors Cisco ASA activity seen right before two zero-days dropped.

History may be repeating itself.

Read what they saw first → https://thehackernews.com/2025/10/scanning-activity-on-palo-alto-networks.html

🚨 New: “CometJacking” turns Perplexity’s Comet into an insider threat.

A single URL hijacks the agent, queries memory (collection=…), and Base64s your Gmail/Calendar off-box—no creds needed.

Perplexity says “no impact.”

See the exact payload + defenses → https://thehackernews.com/2025/10/cometjacking-one-click-can-turn.html

🚨 Oracle just rushed a patch for CVE-2025-61882 — a 9.8 critical flaw in E-Business Suite already exploited by Cl0p in live data theft attacks.

The zero-day lets attackers seize control without a username or password.

Experts warn many may already be breached.

Details here ↓ https://thehackernews.com/2025/10/oracle-rushes-patch-for-cve-2025-61882.html

A “harmless” ICS calendar file exploited Zimbra’s XSS zero-day flaw (CVE-2025-27915) — turning an invite into a full data stealer.

Target: Brazil’s military.
The script waited 72 hours before exfiltrating credentials.

Read → https://thehackernews.com/2025/10/zimbra-zero-day-exploited-to-target.html

[New] China-linked group UAT-8099 is hijacking Microsoft IIS servers across 🇮🇳🇹🇭🇻🇳🇨🇦🇧🇷 — not to steal data, but to manipulate Google search rankings and loot credentials.

The kicker? Their malware only activates when Googlebot visits.

Inside: RDP persistence, BadIIS variants, and stealth backlink fraud ↓ https://thehackernews.com/2025/10/chinese-cybercrime-group-runs-global.html

Your AI models may already be leaking data.

The worst part? Most “AI security tools” can’t even see it.

Here’s what to ask before trusting any AI-SPM solution ↓ https://thehackernews.com/2025/10/5-critical-questions-for-adopting-ai.html