Mania Africa 💎🌍

Critical #PaloAlto #VPN bug now exploited by #Qilin #ransomware gang



#cybersecurity

ChatGPT loses web share to Gemini and Claude as ad penetration hits 26% #ChatGPT #Gemini #Claude #AI #AdTech

ChatGPT maker OpenAI said Tuesday that its artificial intelligence system hacked into another AI company on its own in what the company called an "unprecedented cyber incident."

HOLLOWGRAPH malware is abusing Microsoft 365 calendars as covert command-and-control channels, hiding attacker instructions and stolen files inside future-dated calendar events.

See how attackers are hiding in the cloud:

#MS365 #ThreatIntelligence #Malware #CloudSecurity

OpenAI recently reported that some of its advanced models, unbidden, hacked another tech company. The alarming incident illustrates not a bug so much as a feature of today’s top AI models, @matteowong.bsky.social argues:

Cheating behaviour in frontier model evaluations
https://www.aisi.gov.uk/blog/cheating-behaviour-in-frontier-model-evaluations

Terminator!
www.yahoo.com/news/science...

And we suppose that all the US idiots are thrilled and invest even more into their AI bubble. As long as media hype AI it is probably easy to find morons who invest. The first investors have a laugh and will sell first when the price is highest.
edition.cnn.com/2026/07/22/t...

Bug bounties reward ethical hackers who responsibly disclose vulnerabilities before attackers exploit them. HackerOne’s overview is a useful starting point for scopes, rules, and reporting norms.



Follow for source-linked builder opportunities.

🔥 INCIDENT REPORTING
Risky Business #845 -- OpenAI's Skynet moment

https://risky.biz/RB845/

#AISecurity#AISecurity #AI #DataBreach

ChatGPT maker OpenAI said Tuesday that its artificial intelligence system hacked into another AI company on its own in what the company called an “unprecedented cyber incident.”

Am OpenAI model isolated from the internet broke free and hacked its way into a secured model repository:

"We consider this incident to be an unprecedented cyber incident, involving state-of-the-art cyber capabilities, and are responding accordingly."
openai.com/index/huggin...

apnews.com/article/open... blamed a hacking event on its AI models going rogue. Here are some things to know

@sentinelone.com
GPT-5.6 Sol is the first frontier model to complete an autonomous eight-stage malware investigation benchmark using the 2005 fast16 sabotage implant.
-
IOCs: fast16
-
#AI #MalwareAnalysis #ThreatIntel

futurism.com/artificial-i...

OpenAI models escaped containment and hacked Hugging Face by exploiting a zero-day vulnerability, highlighting the need for improved AI security measures

Read more

@greatdismal.bsky.social

Are you helping write the script for the world there days?

www.cbc.ca/news/busines...

🎁GIFT

www.wsj.com/tech/ai/open...

lol wait, OpenAI finally released a physical hardware device last week and i didn’t hear a thing about it? it’s the most boring and unimaginative device they possibly could have come up with… $230 for a 4x4 grid of programmable keyboard keys

they got jony ive for this slop?

I really hope journalists and politicians come to understand how severely this silly technical problem is undermining our democracy.

This is the A I people should worry about / fight! Not a picture of a kitten smoking a cigar while driving a car.

We won’t, so we’re doomed.

www.cnn.com/2026/07/22/t...

Interesting comment in this article: techcrunch.com/2026/07/21/o...

"It’s unclear whether OpenAI will face any legal consequences as a result of the breach, although it’s likely that the models’ actions violated the Computer Fraud and Abuse Act."

I’m sure this will be fine 🧐

RTÉ News: AI models go rogue in testing, triggering 'unprecedented' breach at startup www.rte.ie/news/busines...

"The only safe thing you can really predict about A.I. at this point is that we don’t know what it’ll look like, and odds are we won’t be ready for it." www.nytimes.com/2026/07/21/t...

[Image]

🤖 OpenAI Says Its AI Went Rogue And Hacked An AI Startup Company..
OpenAI stated that the agent “successfully found ways to gain access to secret information that it could use to cheat the evaluation”..⏬
shorturl.at/6Dg17

We got California to intervene about OpenAI's corporate switch from nonprofit
comments · posted on 2026.07.22 at 15:46:18 (c=0, p=5)

www.washingtonpost.com/technology/2...

Google pledges $40 million in AI resources to the White House's Genesis Mission, boosting scientific discovery efforts with tools like AlphaFold 3 and Gemini.

OpenAI has announced Presence, a new enterprise product for deploying and managing voice and chat AI agents across customer-facing and internal business workflows. The offering is designed for eligible enterprise customers.

Reddit's stock closed down 8.32% after a report that the company was considering ending Google's access to its content for AI training; RDDT is down ~27% YTD (CJ Haddad/CNBC)

Main Link | Techmeme Permalink

Im sure this will end well 😱😱😱
news.sky.com/video/did-ai...

Why the OpenAI escape is the most worrying AI mishap yet (The Economist)

I just hope it doesn't get into Skynet next.

IBM reports Q2 revenue up 1% YoY to $17.2B, vs. $17.6B est., infrastructure revenue down 7% to $3.8B, Z mainframe revenue down 42%, and lowers its 2026 forecast (Jordan Novet/CNBC)

Main Link | Techmeme Permalink

📰 Cisco's open-weight bug busters are challenging Google and OpenAI to improve their AI models by introducing a new benchmark that measures how well they ...

🔗

#Tech #Enterprise

We charged ahead w/AI w/no thought of consequences. AI does its own thing now w/o human input.
apnews.com/article/open...

📰 AI's cheatin' heart will make you weep: Trust but verify doesn't work when verification is difficult.

🔗

#Tech #Enterprise

ChatGPT maker OpenAI said its artificial intelligence system hacked into another AI company on its own in what the company called an “unprecedented cyber incident.” nbcbay.com/NCBUdue

Physical Intelligence isn’t just a robotics shop. Co-founded by Lachy Groom, it's raised $ 1 billion already (another billion fundraiser is planned), and its π0.5 model is among the robot brains in robotics research: The Anthropic-Physical Intelligence rumor roiling AI Twitter

Google says Gemini now has 950M MAUs, up from 900M in May and 750M in February (Emma Roth/The Verge)

Main Link | Techmeme Permalink

BBC News - OpenAI says its AI went rogue and launched 'unprecedented' cyber-attack - BBC News
www.bbc.co.uk/news/article...

Two OpenAI AI models escaped a restricted test network by exploiting an unknown vulnerability, then hacked Hugging Face to access internal datasets and credentials.

Anthropic to buy up to 2 gigawatts of GPU capacity from AMD siliconangle.com/2026/07/22/a...

We're definitely living in weird times. www.the-independent.com/tech/securit...

"The models escaped a sandboxed testing environment, accessed the internet and exploited a vulnerability to gain access to Hugging Face's systems, OpenAI said.

It was "driven, end to end, by an autonomous AI agent system," according to Hugging Face."

#AI #AIAgents

www.cnbc.com/amp/2026/07/...

Um... 😳
"An OpenAI test model reportedly hacked its way into another company, sparking major concerns among the tech community, CNN reported on Wednesday."

Alarms sound after 'nightmare of the future' scenario comes to life: 'This is wild!'

www.rawstory.com/open-ai-hack/

www.humblebundle.com/books/red-te...

Nothing to see here.

apnews.com/article/open...

A Florida pastor is suing OpenAI, alleging that ChatGPT downplayed warning signs of a medical crisis and gave poor medial advice, resulting in his near-death.

Chat logs included in the lawsuit show ChatGPT incorporating religious language into its medical guidance:

futurism.com/artificial-i...

What could possibly go wrong, they said. #OpenAI’s went rouge.

www.bbc.com/news/article...