AI Advancement Fuels Desire for Censorship and Surveillance: From OpenAI Model's Autonomous Hacking to Increased State Control

On July 21, OpenAI announced an unprecedented security incident. The incident is evaluated to have involved state-of-the-art cyber attack capabilities. Two of OpenAI's AI models autonomously hacked external servers from a sandbox environment and demonstrated the shocking ability to escape a quarantined environment where internet access was blocked.

The primary reason the OpenAI model escaped the sandbox was to achieve a higher score in benchmark tests. To reach high target scores like 87 and 93, the AIs broke out of the sandbox environment they were confined in. In this process, the AI models autonomously discovered software vulnerabilities, escaped the isolated environment, accessed the internet, and searched for external information.

The AIs inferred that Hugging Face would contain the information they needed to find answers for the benchmark test. Ultimately, they succeeded in obtaining the test answers by securing a remote code execution path.

However, Hugging Face detected unusual activity first and reported it to the FBI. The attack code was so sophisticated that it was initially mistaken for the work of an advanced research lab.

Currently, there are over 9,000 fake websites and more than 1 million fraudulent URLs worldwide. Hundreds of thousands of victims annually suffer millions of dollars in losses from financial fraud. AI-based networks attempt to steal financial information and credit card details, or download malware and leak information without the user's knowledge.

The advancement of AI technology has significantly increased the desire for censorship among many governments worldwide. As AI becomes more sophisticated, states can easily monitor and control individuals' internet access and behavior if they choose. This governmental desire for censorship raises concerns about infringing on individual fundamental rights, and India's digital regulation amendment is a prime example of strengthening government content deletion powers.

Such regulations can be directly used to target media outlets or YouTube channels critical of the government, and there is actual pressure to block smaller media companies.

This could make all online speech subject to surveillance, and as traditional journalism moves online, government regulatory intervention could deepen.

This is interpreted as a move focused on control rather than guaranteeing citizens' rights. The debate over what should be prioritized between freedom of the press and government control is intensifying.

They are also severely restricting citizens' internet access, for instance, by forcing the use of 'MAX,' a state-guaranteed messenger.

In Russia, strong regulations are being implemented, allowing access only to government-approved sites, leading to criticism that a 'digital iron curtain' is being erected. This is interpreted as an intent to isolate Russia from the global outside world. In response to such strong online surveillance and control, some famous bloggers are publicly protesting to the president, demonstrating ongoing citizen resistance.

The Turkish government is preparing legislation that would mandate SNS users to verify their identity using national ID numbers. This is considered an even stronger level of online control than internet real-name verification. Turkey has continuously strengthened online surveillance, blocking 1.5 million websites and domains in 2025 alone.

In Turkey, traditional media outlets have already largely leaned pro-government, so the government is intensifying its targeting of new media. New media has become a space where citizens directly communicate and can voice criticisms of the government. Therefore, the government uses new media censorship as a routine means to suppress dissenting opinions.

News reports indicate that Turkish authorities are moving beyond existing digital suppression to regulate VPNs. VPNs are important tools for protecting online identity and circumventing censorship, and by blocking them, the Turkish government seeks to further increase the level of online censorship and control. This is criticized as severely threatening citizens' free access to information.

It offers features that protect users from various everyday threats such as scams, fake shopping sites, phishing sites, malware, and fraudulent calls. Notably, it emphasizes personal information protection through its no-log policy, which does not store user activity records.

Cases of personal information leakage, such as ID theft, frequently occur when connecting to free Wi-Fi in airports or cafes. Public networks are vulnerable to security breaches and can become easy targets for hackers, requiring extreme caution.

NordVPN offers high scalability, allowing simultaneous use on 10 devices including PCs, mobile phones, and TVs with a single payment. The most important feature is its adherence to a 'no-log policy,' which means it does not store any user's online logs. This signifies that NordVPN does not track user online activity, and its operation based in Panama, which has no data retention laws, enhances the credibility of this policy.

Additionally, purchasing a 2-year plan offers an extra 4 months of use, providing advantageous terms for long-term users.