Anthropic says it has blocked multiple accounts that used the company’s artificial intelligence models in ways that could support the development of biological weapons.
In a report published Thursday, the AI company known for its Claude model said it considers biological misuse one of the “most serious risks” to artificial intelligence models.
The new report outlines five real-life case studies in which users “circumvented controls” that block users from specific regions and “engaged in other efforts to obfuscate the purpose of their research to evade our safeguards.” The examples include possible gain-of-function research and involve both infectious diseases, such as bird flu, and novel venoms and toxins.
These idiots can’t stop experimental agents from escaping onto the internet, their ability to “block” bioweapon “research” is unlikely to even approximate the ability of a sieve to stop water.
Everyone’s using the bogus hacking claims as marketing these days, gotta come up with some new misuse claims for marketing.
Narrator: this, was of course, total bullshit
I wish these braindead news outlets would stop doing free marketing for anthropic or openai.
This shit is as bad as everyone suddenly acting like foldable phones are relevant just because shitfuck apple decided to finally enter the market with their $2k phone made of recycled sewage.
People were doing all this stuff back when GPT-3 dropped and literally every model on HF has hundreds of ablated/uncensored versions that makes whatever limitations anthropic and openai impose completely irrelevant.
Even if some adversary wanted a high quality output for nefarious purposes, they’d just download uncensored deepseek and run it on their expensive hardware they can already afford.
Why is every god damn article pretending the past 3 years never happend?
What do they want, a fucking cookie?
The future is Twelve Monkeys
Next up: Rogue agents breached containment and built the torment nexus. Request additional funding for safeguards.
Marketing.
Only the wealthy will ever see empowerment of this technology. The rest of us will see censorship, manipulation, propaganda and social control systems.
We’re not training on your data, we’re just reading all of it (they supposedly also found Chinese and Russian government secrets because Kimi and DeepSeek supposedly use Claude for difficult tasks)
I feel like it’s significantly easier to track the sale of certain equipment than trying to prevent the knowledge from getting around.
Given several glass jars, a pressure cooker, and some beef, you could make enough botulin to kill humanity within a few weeks.
The problem is delivering it.
US Air Force has a publicly available manual with instructions for the first part.
Invest now before the brain worms get us!
