Open-source AI is more dangerous than closed-source AI

For models with dangerous capabilities – including highly cyber-capable models – open weight release therefore creates a persistent and irreversible risk of misuse.

This unlocks a new avenue for platforms that host open-source models and for law enforcement to actually test whether a model is capable of generating CSAM. Before, we had no way of measuring this. It was a huge blind spot that some people were takin...

Because open-weight safeguards can often be removed, public release of such models may make it impossible to prevent a wide range of actors from accessing dangerous capabilities.

Open-weight AI models have a sovereignty advantage in that sensitive data can remain local and the AI producer cannot revoke access. The flip side is that fine-tuning can also degrade or remove safeguards against misuse. Producers of open-weight AI m...

Open-weight models – models of which the core parameters are made available for others to run and adapt – are an important medium-term issue. Evidence suggests they may lag the most advanced closed models by only four to eight months, and that their ...

You can’t patch a released, open-weight model. This is the property that breaks the assumptions of the prior model of responsible disclosure. Open-source software can be patched in place; a deployer updates and the hazard closes. An open-weight model...

When a foreign adversary copies an American model, strips those safeguards out, and releases it to the world, those protections vanish.

The 71pp open-weight span is therefore not an artifact of quantization or scale: the same low/high codification grouping appears across model families and deployment surfaces, attenuated in absolute level (highest closed-model cell 40.4% vs. open-wei...

Everybody can download and operate their own state-of-the-art model and use it for great things and terrible things.

In this paper, we show that open-weight safeguards are susceptible to simpler strategies that, despite being well known, have not been systematically evaluated against these safeguards. Specifically, we evaluate two low-cost attacks--abliteration and...

Open-source AI models, which can be accessed, modified, and distributed freely and pseudonymously on the Internet, are also largely untethered to any ethical or enforceable commitment not to be used for harmful purposes.

The adoption of CyberStrikeAI is poised to accelerate, representing a concerning evolution in the proliferation of AI-augmented offensive security tools.

But open sourcing, that's just sheer catastrophe. [...] wrong approach, the wrong ideal. [...] where if you could align it it would take time [...] not kill everyone.

When you open source a big model and give people the weights [...] You get people using that open source model and training it, refining it to do something bad like cyberattacks or making biological weapons because it's much easier to refine the mode...

Open-Source AI Is Uniquely Dangerous [...] As things stand today, unsecured AI poses an enormous risk that we are not yet able to contain.

Nefarious actors can use open models maliciously, [...] whereas developers of closed models have more means to identify and disrupt this malicious use.

You basically have a bomb that you're making available for free, and you don’t have any way to defuse it necessarily. It’s just an obviously fallacious argument. We didn’t do that with nuclear weapons: we didn’t say ‘the way to protect the world fro...

Because they can also be run more cheaply than proprietary models, the point at which open-weight models become dangerous and practical to use may arrive before they fully catch up with frontier models.

As open-weight models approach the kind of cyber and biosecurity risks of Mythos and other leading-edge models, China may make the same calculation as the US and find them to be too dangerous to be released, especially in open form.

Open-weight releases such as Kimi K2.5 are already benchmarked on par with frontier closed models on cybersecurity tasks. Safeguards on open-weight models can also be removed through fine-tuning, a well-established procedure.

An attacker can run it locally without safety guardrails, fine-tune it against their specific targets, and operate with zero visibility to any provider or defender.

Hackers are already talking in Russian-language forums about how easy it is to jailbreak GLM-5.2 for hacking tasks.

Muse Spark had triggered safety checks that made it "not suitable for open sourcing."

Whilst safeguards applied by responsible model developers can limit misuse of AI, these protections can often be bypassed (and in open‑weight models can be removed or are absent from the start). Publicly-available AI model weights can be modified, an...

endowing rogue nations or terrorists with tools to synthesize a deadly virus. [...] keep the “weights” of the most powerful models out of the public’s hands.

It is unquestionably both. We can use the frontier models but also open weight models to prove a solution which gives defenders powerful capabilities and on their terms.

These studies, on their own, are insufficient evidence to demonstrate increased marginal societal risk from open foundation models.

When a frontier model is attacking you and moving laterally inside your infrastructure, defenders need wide access to near-frontier tools within hours or even minutes, rather than being pointed toward a closed-door platform.

These Chinese models are excellent. Open-source models that are excellent should be used.

Xi urged countries to seize the “historic opportunity” of open-source AI.

Open-source AI is more than a catalyst for commercial innovation. It is a strategic advantage for the U.S. military, enabling American companies to push the technology forward while giving the DOD a foundation it can inspect, secure, adapt and deploy...

Open-source AI is a shared asset for all humanity.

Building global trust in advanced AI will require more than shared principles, it will require shared, practical standards. The Appia Foundation can help translate emerging frontier AI practices into open specifications and technical standards that p...

Study of 34 AI models from 10 global providers finds open-source AI models are not less safe, reasoning models are hardest to exploit and smaller models are the most vulnerable.

The governance of open-weight artificial intelligence (AI) models has been framed as a binary choice: openness as risk, restriction as safety. This paper challenges that framing, arguing that access restrictions, without governed alternatives, may di...

Transparency is the greatest safeguard. [...] When codes are open and shared, millions of young minds can make them safer and better.

Open source does not eliminate risk. It changes how risk is managed. It allows more researchers, developers, and defenders to examine systems, test assumptions, surface weaknesses, and harden code under real-world conditions. [...] If frontier models...

Open-source AI allows the public to audit and verify algorithms, [...] It is also more secure because any attacks or exploits can be identified and remediated.

From a business perspective, the difference between open and closed is a little bit overblown. From a security perspective, the difference between open and closed models is, for some intents and purposes, overblown. The most important thing is how po...

But the real danger lies in centralized, black-box systems that resist open auditing.

America’s footprint in artificial intelligence is prodigious, and it is hard to overstate how consequential this is for the American national interest if it further develops with the right balance between innovation and guardrails. Into this new tech...

The researchers found that available evidence does not show that open models are riskier than closed models or information already available [...] like online searching.

So I don't see any risk associated with open sourcing models. I only see benefits.

Lawmakers in Europe and California are worried about that open-source AI is “dangerous.” On the contrary — there is nothing dangerous about transparency.

I see the danger of this concentration of power through proprietary AI systems as a much bigger danger than everything else. What works against this is people who think that for reasons of security, we should keep AI systems under lock and key becaus...

[...] an open innovation ecosystem and see the immense value of community-built technology as the best way to grow AI responsibly and mitigate potential misuse.

Since the launch of powerful open models like the Llama, Falcon, Mistral, and Stable Diffusion families, critics have pressed to keep other such genies in the bottle. "Open source software and open data can be an extraordinary resource for furthering...

Are open-source AI models more dangerous than closed models like GPT-4? A new study says no, and offers recommendations for policymakers. Open Foundation Models (OFMs) offer significant benefits by fostering competition, accelerating innovation, and...

open source is not more dangerous than closed source

Open source provides an opportunity for many people to examine those technologies, to test them in a variety of settings and to provide fixes [...]

SB 1047 will stifle open source AI and hinder AI innovation [...] and therefore will make AI less safe.

Open-source AI fosters innovation, improves security, and lowers costs—critical benefits for businesses, researchers, and everyday users around the world.