are panicking. Many dormant security issues could be found using new
AI cybersecurity models such as Anthropic Mythos, OpenAI 5.4 Cyber or
other similar models. Those models have discovered over 270 security
flaws in widely used open source software.
One tech community from Africa, cyberstorm.mu, got access to a
restricted AI cybersecurity model. They are already using it to find
security flaws in popular open source software such as OpenSSH. One of
their security fixes has already been reviewed and merged.
Cyberstorm.mu is not a newcomer to Computer Security. They have spent
close to 10 years working on improving free and open source software
via hackathons events. This time, they are warning the government of
Mauritius to take the threat of capable AI models seriously.
They are writing a petition addressed to the government of Mauritius
to push for more spending on improving security of Critical
Infrastructure in line with what Singapore, Japan and the USA are
already doing.
Readers are welcome to sign here:
petition.srv1561000.hstgr.cloud/?t=73UnDjcXFNGyXDbycuOhs0Gb
Views: 138