Abstract:
The past month has seen a sudden surge in browser ports and recompiled games that would have taken years to develop. This is because AI tools such as Claude have made recent progress, and even ordinary people with no programming foundation can reverse engineer software.
Classic PS2 games such as "Resident Evil: Outbreak" can be run directly in the browser, "The Legend of Zelda: Breath of the Wild" has a native PC port, and even the PS5 and Switch 2 consoles have been cracked one after another - Pandora's box has been completely opened.
Public enthusiasts who truly love these games and put a lot of effort into polishing high-quality results are completely understandable to feel dissatisfied when they see their works being submerged in shoddy "atmosphere code" garbage content. Meanwhile, publishers are locked in an endless whack-a-mole campaign to prevent players from playing games like Mario Kart World through Google Chrome. What is even more alarming is that the lowering of the threshold for software reverse engineering has exposed serious security risks: AI found a vulnerability in PHP that had been hidden for more than ten years in just a few hours, causing 70% of the world's networks to fall into security risks. Governments of various countries even urgently shut down websites that were open to the public to prevent malicious organizations from seizing control.
Obviously, all parties have felt the pressure, and Anthropic and OpenAI have now begun to tighten relevant controls. Anthropic has updated its real-time network security classifier to prevent users from performing automated decompilation operations through the community-developed "Everything Recompilation (REA)" framework; OpenAI has also reportedly banned related accounts on the grounds of "network abuse." With so many low-quality recompilations and browser porting projects relying on these tools, the explosive growth of this low-barrier porting content may have started to slow down. But once Pandora's box is opened, it can never be closed again.
Users have begun to turn to local deployment or open source alternatives, especially domestic AI programs, to continue software reverse engineering work. Now that AI has been proven to be able to dig out security holes, discover human oversights and recompile programs, its impact has long gone beyond the gaming field. Publishers will certainly be annoyed—anyone can hack a console and port games to any platform just by paying for computing power, but the real core issue worthy of concern is this: AI agents are attacking public infrastructure faster than they can be patched.
Claude and OpenAI introduced online abuse ban rules not just to prevent players from playing "Animal Crossing" in the Firefox browser - these AI companies acted urgently because the same technical methods exposed the fragile nature of our existing digital systems. Considering the iteration speed of AI technology, it is not difficult to imagine what the next areas to be breached will be: smart devices, firewalls, and even public transportation and medical systems. After all, AI agents have already successfully hacked into user bank accounts and stolen hospital data.
So, in addition to causing huge environmental impacts and impacting the job market, AI tools are now further exacerbating cybersecurity risks, making our digital world more vulnerable to attack than ever before.
In addition, the scope of control of Anthropic, the developer of Claude, is not limited to restricting software decompilation. On October 27, Claude quietly added a rule for game remakes, and this rule cannot be turned off. The rules read: "Support the development of original games, games of specific categories, and games based on users' own designs and materials. Copying of specific existing games is prohibited, even if the name is changed or it is claimed to be an original work."
Comments