Commenters question whether the rule has teeth
3 Oct 31 3:30 AM · 12d ago · 1 post · 6 comments · 2 sources · development 3 of 3
Discussion on Reddit and elsewhere focused on whether the instruction is easily bypassed by rephrasing and whether it creates any real legal consequence, with some noting open-weight models remain unrestricted.
“We added two sentences to Claude Code's prompt on Opus 5.5 to steer it away from recreating someone else's specific game because using Claude to infringe IP is against our usage policy.”
Addy Osmani, Anthropic member of technical staff · mastodon ↗Addy Osmani Anthropic member of technical staffAnthropic Developer of Claude and Claude CodeChasm X user who discovered the ruleDigital Sun Games Indie game studio
The whole story articlespostscomments the bright band is this development · numbered dots are the others · click one to jump
Reported in the same hours no headline names this development itself — these 1 claim were published in its stretch
-
first by Mastodon, 12d ago · also Tom's Hardware, Mashable
3 more headlines
- Anthropic's Claude AI blocks video game cloning and reverse-engineering Tom's Hardware · 12d ago
- Anthropic takes steps to block Claude from reverse-engineering video games Mashable · 12d ago
- Anthropic tells Claude to stop remaking video games Mastodon · 11d ago
What people said 24 voices · best of 44 · verbatim
-
> preemptively protecting themselves from potential lawsuits ... for offering the means to reverse-engineer their software.I think the legal concern isn't so much the act of reverse engineering itself which is mostly legal (except RE specifically to remove anti-copying protection under DMCA), it's that damages from contributory infringement are…
-
I did indeed! Yay, another English/Lit tech refugee - there are dozens of us! Dozens! 😂💕 There's a monstrous gap between folks who can effectively communicate and those who can't when it comes to AI, and life in general lol +1 +1 +1. Truer words have never been spoken. It's the one saving grace that has allowed us to pivot our…
-
H
Hope you have fun using Claude to reverse and clone, b/c you ain't doing it anymore. https://www. tomshardware.com/tech-industry /artificial-intelligence/anthropics-ai-model-is-now-blocking-attempts-at-video-game-recreation-and-cloning-users-also-claim-non-malicious-generic-reverse-engineering-tasks-are-blocked /insert…
-
They never expected their models could be used for reverse engineering.That's the truth.Now, they're panicking. And this is why the other dude, OpenAI, came out to say that people shouldn't use RAM locally, because they can't control the software anymore.Open weight models, ablated models are going to become the norm, hopefully, locally run.So…
-
Yes the hardware costs a chunk as a one time investment, but mostly it’s slower because you’ll never have the same high-end chips. People probably have no concept of what kind of gpu time a typical inference uses, and how much time it takes to load a model into memory, it just seems mostly instant because we’re basically getting free compute at…
-
The collateral damage here is killing some truly wonderful (and completely legal) retro game preservation decomps, ports, deep mods, community translations/expansions, and academic study of classic titles which Claude has been instrumental in enabling.The retro preservation community doesn't distribute or even link to original game binaries or…
-
That's not entirely fair to Disney (can't believe I wrote that). Their versions and their characterizations of the public domain characters are what they own, and they should. That is their creative labour. Also their specific versions of the stories. What they don' own is the story and alternate characterizations. People really don't understand…
-
Hmm - I'm implementing a browser-based version of wolfpack empire from my college days a few decades back [1], and claude is happily helping. Perhaps it doesn't care about open-source games.1:
-
Another double standard - if I blatantly copied a game and sold it for money on demand, I'll likely get sued by the publisher. But when Claude blatantly copies a game and sells it for tokens on demand, somehow Claude is immune from any litigation? So why are AI companies are allowed to profit from distributing and selling copies and ripoffs of…
-
how would they know im reversing a game if i let it analyse a bunch of decompiled results? aslong as u'd cut out obvious identifiers like perhaps urls / connection info which u dont really need to reverse the game it will be very tedious.id expect this effort to be either inefficient or yield lots of false positives??
-
Pretty much, we have claude 5.5 at work and despite what the AI bros online will say you still need to hand hold and rigorously check the output to the point where I tend to just do fixes myself since it's faster. (incoming YoU jUsT nEeD bEtTeR pRoMpTs BrO) The timebomb of newer engineers not building the skills and outsourcing to the LLM is also…
-
> we're going to police what kind of programs you can compile" ?We don't have to imagine. Oracle's Java said you weren't allowed to use it in nuclear power plants. back in 1998. Naturally the number of people that were held back by that was a bit smaller, but the idea's not without precedence.
-
It’s how many of these tech companies operate, Netflix, Uber, etc they subsidize their product, operate at a loss for years living on VC money, then get profitable by raising their prices over time. Also the entire industry is in it’s infancy and likely in a bubble right now, if that bubble bursts then we might start to see consolidation or…
-
This will work for approximately a year until ablated open weights models catch up, and then it will be open season. My 2c is that we need to figure out reputation systems that are resilient to low effort cloning rather than try to prevent it, because the latter approach is a non-starter outside of a very short time horizon.
-
If I had to take a stab, I think the overseer feeds the model a new prompt to output. So it goes into the overseer, overseer says "no", rewrites the prompt to be safe, hands it back to the model, and the model pushes the new prompt. At least that's what I "feel" is happening. It's minute changes in the response time and changes to the sentence…
-
So they used tons of information - books, arts, code, articles and so on all over the internet, some without consent or license (case with book publishers) but when it comes to regular people using it it’s now problem?Law and accountability is for poor people apparently, power is for rich.
-
Yeah nice, I’ve just been trying to work out if I’m better using a combo of Claude/Deepseek/etc vs buying hardware for what I want to do at home, I’ve been pretty lucky at work as I have a massive budget to just run Claude with with MCP connections to Ida/Windbg etc and not have to think much about choosing a specific model and setting it all up…
-
This is why we need open AI (not OpenAI). Some duopoly deciding what goes on a whim is not acceptableI can sorta see them suppressing instructions for say bioweapons (that seems fair to me), but this demonstrates that they'll bend for corporate interests.
-
what you're hoping for and postulating is an abliterated model of something like a new qwen let me ask you, are you looking at the gguf file and inspecting the jinja? because if you're not, odds are that you're going to download something malicious and get blindsided later. and even if you're doing that it's still possible to be attacked in the…
-
remember when llms were a tool and you could use them however you wanted? pepperidge farms rememberscan't wait for the day when you end up in jail for the wrong prompt or when you will be served ads through the kazoo.
-
Exactly what I was trying to say earlier, their costs are way too high to continue at these price points indefinitely, at some point prices are going to rise significantly for those customers (eg enterprises) that can’t just quit AI. This will eventually push them towards cheaper open weight/source models
-
I wonder if this is the start of a chain of events that will lead to anything that requires more tech savvy than editing an Excel spreadsheet to be limited to enterprise customers, while consumers get a dumbed-down chat bot.
-
System prompts are easily bypassed with clever rephrasing and rewording things around it. There's no actual legal liability if people bypass the system prompt filters and still are able to get Claude to spit out copyright material.
-
There is some irony there. So asking the AI to mod or reverse-engineer an existing game is a no-go, but asking it to create a "new" game using stolen assets and stolen code is fine...
All 3 developments of Anthropic confirms Claude Code rule blocking game… →
MastodonNewswiresGoogle NewsHacker NewsLobstersReddit