The late NBA player Kobe Bryant appeared in the music video for Kendrick Lamar's song "The Heart Part 5", which stunned the audience. Deepfake technology was employed in the video to pay tribute to the late legend.
Deepfakes are images and videos that have been altered with advanced deep learning technologies such as autoencoders or generative adversarial networks.
With the support of deepfake technology, realistic yet manipulated media assets can be easily generated. However, deepfake technology is deceptive. The technology is utilised in virtual reality, video games, and filmmaking, but it might also be used as a weapon in cyberwarfare, the fifth dimension of warfare. Additionally, it can be used to share false information to influence public opinion along with political agendas.
Cybercrime is on the rise as the internet's global penetration grows. According to the National Crime Records Bureau, there were around 50,000 incidents of cybercrime in 2020. The national capital witnessed a 111% increase in cybercrime in 2021 compared to 2020 as reported by NCRB.
The majority of these incidents involved online fraud, online sexual harassment, and the release of private content, among other things. Deepfake technology may lead to an increase in such incidents that are weaponized for financial gain.
Notably, the technology is not only a threat to the right to privacy protected by Article 21 of the Constitution, but it also plays a key role in cases of humiliation, misinformation, and defamation. Whaling attacks, deepfake voice phishing, and other frauds that target individuals and companies are thus likely to rise.
Mitigation Tips
The difficulties caused by deepfakes can be addressed using ChatGPT, the generative AI that has recently gained attention. To offer viable options, ChatGPT can be integrated into search engines. In order to combat the dissemination of misinformation, the AI-enabled ChatGPT, based on Natural Language Processing, is trained to reject inappropriate requests. It can also process complicated algorithms to carry out complex reasoning operations.
In order to swiftly purge such information from the internet after deployment, the dataset needs to be fine-tuned using supervised learning. It can be further tweaked due to its accessibility to offer a quicker, more practical solution that is also affordable. However, to stop AI from scooping up new deepfakes from the test set, the train set must be constantly monitored.
Additionally, a greater influx of cyber security specialists is required to achieve this. India's GDP currently only accounts for 0.7% of research and development, compared to 3.3% in affluent nations like the United States of America. The National Cyber Security Policy of 2013 must be improved in order to adapt to new technologies and stop the spread of cybercrimes as these manipulations become more complex over time.
