
Car traffic on the Crimean Bridge restored in reverse mode
Russian Deputy Prime Minister Khusnullin announced the return of car traffic to the Crimean Bridge, noting that not all lanes are open to vehicles.

Russian Deputy Prime Minister Khusnullin announced the return of car traffic to the Crimean Bridge, noting that not all lanes are open to vehicles.

According to Google, hackers who used Russian in technical documentation created AI agents that likely independently found vulnerabilities and attacked French websites without human intervention.

The British star posed next to statues of previous 007 actors, which is seen as a hint at a possible appointment.

Ukraine's president announced online consultations with the US and the EU. The parties agreed on a next meeting in the German capital.

Chupa Chups has launched a line of lollipops containing miniature DreamWorks toys inside.

Ukrainian President stated that Moscow is increasing its own missile weapons production and receiving ballistic missiles from North Korea, posing a threat to regional security.

The relevant authority confirmed the attack by a Russian drone on a gas station in Dnipropetrovsk region. No casualties, damage assessment ongoing.

Open sources published information on the MASON structure and its head, Daniel Glurjidze-Menshov.

The cyber intelligence division of tech giant Google reported a potentially first-in-history fully automated cyberattack carried out using artificial intelligence. This became known from a company statement published by Politico.
Google experts noted that hackers who used Russian in their technical documentation deployed a series of AI-based agents. According to the company's data, these systems "continuously compared code before and after fixes, identified vulnerabilities, and automatically developed exploits – without human involvement".
This means that artificial intelligence independently identified software flaws and developed methods to exploit them. Previously, even the most advanced AI models could not perform such actions fully autonomously, often requiring the assistance of skilled hackers to correct actions or fix errors.
According to Google, the perpetrators attacked French websites operating in the media, entertainment, and education sectors. The main change in the hackers' approach is that AI models allow attacks to be accelerated and made more targeted.
Politico notes that this development could be a turning point in the use of artificial intelligence for cyberattacks. The topic of threat automation has concerned politicians since Anthropic released its specialised Mythos model in April.
Such a development will likely alarm governments and critical infrastructure operators. This is happening against the backdrop of the recent discovery of serious vulnerabilities in the PHP programming language, which powers 70% of the web, as well as a series of hacks of supermarket refrigeration equipment, for which Iran is partially held responsible.
Comments
No comments yet. Be the first!