AI Brain-Reading Tech Emerges: Who Will Protect the Last Frontier of Privacy?

Israeli researchers have unveiled a new AI model called 'Brain-IT' capable of reconstructing visual scenes solely from brain signals. Developed by Professor Michal Iramani's team at the Weizmann Institute of Science, the system analyzes fMRI data to recreate images viewed by participants, with accuracy surpassing previous methods.
The research involved 8 participants viewing thousands of images while their brain activity was recorded, generating approximately 73,000 data points. Unlike earlier models that struggled with color and composition, Brain-IT accurately reproduces both content and fine details. The system also significantly reduces training time—requiring only about one hour of scan data compared to previous models needing up to 40 hours for similar results.
While current technology is limited to reconstructing what a person is actively viewing under controlled conditions, researchers warn that future applications could extend to interpreting dreams or emotions. However, the team emphasizes that today's capabilities are far from reading thoughts covertly or analyzing brain activity without consent.
The development has intensified debates about 'neuro rights'—a growing movement to legally protect neural data as a fundamental privacy right. Chile amended its constitution in 2021 to safeguard brain activity information, while Colorado and California passed laws in 2024 classifying neural data as sensitive personal information.
Experts caution that just as smartphones and social media eroded expectations of digital privacy, neural data could become a new battleground for personal autonomy. Unlike passwords or credit cards, neural data cannot be easily reset if compromised. The article stresses the need for clear legal frameworks to govern how neural data is collected, stored, and used—particularly in contexts like employment, insurance, and law enforcement.
Blockchain technology has been suggested as a potential tool for managing consent and access logs, though researchers warn that storing raw brain signals on public ledgers would pose significant privacy risks. The core principle remains: individuals must retain control over their neural data, including the right to refuse unwanted scans or prevent discrimination based on mental states.
As AI advances beyond language processing into sensory interpretation, the question of whether thoughts can remain private becomes increasingly urgent. The article concludes that rules protecting this final frontier of privacy should be established before the technology fully matures—not after it has already been widely adopted.
Korean Source
This article is an English localization of a Korean-language crypto news report. Original headline: [권성민 칼럼] AI가 뇌를 읽는 시대…인간의 마지막 사생활은 누가 지키나