- JPMorgan CEO Jamie Dimon warns that checks may be necessary for Claude Mythos
- Anthropic’s AI model is highly advanced and has been shown to detect zero-day vulnerabilities and even develop working exploits.
- The US government has previously ordered Anthropic to block access to foreign citizens, citing security concerns.
Artificial intelligence is becoming increasingly powerful, a fact highlighted by the US government’s recent directive to limit access to Anthropic’s Claude Mythos model.
Now, JPMorgan’s chief executive has described the risks posed by the technology as a “real problem” and compared broad access to AI to “giving ballistic missiles to individuals.”
Speaking at the Pennsylvania Defense and Innovation Summit, JPMorgan’s Jamie Dimon highlighted the risks posed by AI, which has already been shown to identify and exploit cybersecurity challenges.
Mythos is not Skynet
Currently, Claude Mythos is limited to a small selection of federal and military organizations, companies, and departments. Public access to the AI has been blocked and global access has been blocked due to potential security issues with this powerful AI.
Claude Mythos is not the kind of threat that can become self-aware and take over military apparatuses like the fictional Skynet of the terminator film series. However, it has the ability to cause immense damage in the wrong hands.
AI is capable of detecting cybersecurity vulnerabilities and reporting on them; It is also capable of generating automatic exploits. So for white hat cybersecurity analysts, it’s a powerful defensive tool, but in the wrong hands, it’s a destructive power. Additionally, its automated capabilities make it easier for black hats to expand their cybercrime operations and make ransomware, phishing, and other data-driven scams more profitable.
It’s not just in cybersecurity where Anthropic’s Claude Mythos AI has demonstrated such incredible capabilities. Scientific research, specifically biology, can be targeted by attackers who use AI to convert complex concepts into terms, potentially misusing the information.
There is also the challenge of a strategic imbalance between individual nations or agencies having exclusive access to the technology.
Will the public ever be able to access Anthropic’s Claude Mythos?
Given security considerations, it seems unlikely that Claude Mythos will be publicly available any time soon. However, Anthropic has already stated that it intends to make Mythos-level features and capabilities more widely available.
However, for this to happen, various safeguards need to be developed and applied.
The most likely scenario is that public access to Mythos will eventually be unlocked, with restrictions on specific features related to cybersecurity and biological/medical tasks and research.
Project Glasswing is already up and running with a handful of key partners (including Amazon Web Services, Anthropic, Apple, Broadcom, Cisco, CrowdStrike, and others) and will likely expand until the safeguards are proven fit for purpose and public use.
Through PakGazette
Follow TechRadar on Google News and add us as a preferred source to receive news, reviews and opinions from our experts in your feeds.
