For centuries, national security of states has only been threatened by agents and spies. Furthermore, the risk can also come from models artificial intelligence.
Not often does the five-large cybersecurity services happen «Anglo-Saxon» I would like to thank the President-in-Office of the Council and the Commission for their work.
ARTICLE CONTINUES AFTER ADVERSION
This is the alliance. «Five Eyes» [«Five Eyes»), in which the relevant services are involved by USA, Canada, the United Kingdom, Australia and New Zealand. According to a public description of the CIA alliance, the partnership helps member states fill information gaps and strengthen their capabilities against common threats such as terrorism, proliferation of weapons of mass destruction and cyber-hazards.
Artificial intelligence: How much has it affected the Greek labour market?
Anthropic's Nightmarish admission: We're losing control of AI, we have to stop now.
A sound joint statement on AI after Trump's preventive measure
So what did he say? «Five Eyes» Who knocks? That if the necessary measures are not taken, powerful artificial intelligence (AI) models are likely to be able in a few months to overthrow governments and businesses.
ARTICLE CONTINUES AFTER ADVERSION
In a similar report, in fact, Guardian notes that this rare joint statement comes after the government's decision Trump, earlier this month, ban citizens of other countries working in U.S. public services using Anthropic's multi-advertised artificial intelligence model, called Fable.
In other words, since the Trump government has taken a precautionary measure aimed at shielding U.S. national security from potential risks, the country's cybersecurity services along with «allies» They were given a sound warning that may concern more countries and more artificial intelligence models.
The timetable «Not years, but months»
In the statement issued today, Monday, it is noted that while artificial intelligence «will help us improve cyber defense over time, also accelerates the speed, scale and complexity of cyber threats».
«The pioneering artificial intelligence models are expected to overcome the current expectations of the industry, radically transforming both aggressive and defensive cyber-governments. The timetable is not years, but months», stated their warning «Five Eyes». «In this environment, cyber-resistantness is an integral element in promoting business continuity, market confidence and value in the long term».
ARTICLE CONTINUES AFTER ADVERSION
Cybersecurity services said rapid developments in artificial intelligence models show that this technology will reduce obstacles to malicious perpetrators and increase the speed and complexity of attacks. «A reaction is required that will involve all organizations and society as a whole.», marked.
In fact, the services stress that the issue is now a political one. «Cyberhazard can no longer be treated as a purely technical issue. This is a key business risk and responsibility of leadership», emphasize.
Although the statement does not mention nominally specific artificial intelligence models or companies, many around the world have turned their attention to Anthropic's advanced tool series, observes Guardian, referring also to the Trump administration's move.
American lawmakers worry about the potential of Anthropic models
According to the magazine «The Economist», Democratic Senator Mark Warner, vice president of the Senate Intelligence Committee, said on June 11 that General Joshua Rand, director of the National Security Agency (NSA) and the American Cyber Command, told him that the model «Mythos» Anthropic managed to infiltrate into «almost all our classified systems» in a few hours and not weeks.
ARTICLE CONTINUES AFTER ADVERSION
Ωστόσο, πριν από λίγες ώρες προέκυψε πως οι αρχές είχαν εξουσιοδοτήσει το «Mythos» να προσπαθήσει να διεισδύσει προκειμένου να δοκιμαστούν οι αντοχές του συστήματος. Δοκιμάστηκαν και, όπως φαίνεται, βρέθηκαν ανεπαρκείς.
Follow it. on Google News and learn all the news first
See all the last News From Greece and the World,
