DeepFakeLib Demystified: Ethical AI for Crafting Consensual Likenesses & Avoid Potential Lawsuits!
Hey…Hey…Hey….! It seems our lives will be forever intertwined with AI. Our livelihoods and those of our future generations will all be profoundly influenced by it. AI might soon undertake every mundane task, eliminating the need for us to engage in simple decision-making.
For instance, I don’t want to ponder why the house is cold or whether I should turn on the thermostat.
Instead, I imagine a system that senses my body temperature, interprets my need for comfort, and adjusts the environment to keep me cozy. Haha!
Recently, On AI news….
Scarlett Johansson Sues AI App for Unauthorized Use of Her Image in Ad!
Hollywood actress Scarlett Johansson has initiated legal proceedings against an AI application, Lisa AI, for illicitly exploiting her image and voice in their advertising campaign.
The ad, which was disseminated on social media platforms, showcased a digital imitation of Johansson without her consent, sparking legal action from the actress’s camp.
The contentious advertisement, which has now been removed, featured a deepfake version of Johansson promoting the app’s capabilities.
Her attorney, Kevin Yorn, emphasized the seriousness with which they approach such infringements and mentioned that they would seek all available legal remedies.
This event underscores the escalating concerns surrounding the misuse of AI technology in mimicking celebrities. It serves as a stark reminder of the legal boundaries that protect individuals’ rights to their likeness and voice. California’s laws, in particular, provide a strong legal framework for individuals like Johansson to challenge unauthorized commercial use of their identity.
Johansson’s case follows similar incidents, including one involving actor Tom Hanks, who publicly denounced the use of his AI-generated likeness in a dental advertisement. These incidents are drawing attention to the need for stricter regulation and enforcement in the AI domain to prevent such exploitations.
As the legal process unfolds, the case of Scarlett Johansson vs. Lisa AI is expected to be closely monitored as a potential precedent-setter in the realm of AI and rights to one’s image and likeness.
Let’s learn something about AI & Fake identity through example: How someone might programmatically generate a likeness of a person using AI.
The technology for creating deepfakes or synthetic media often uses machine learning models called Generative Adversarial Networks (GANs).
Below is a simplified conceptual explanation of how one might approach this task using Python and a hypothetical AI library.
Note that in practice, the use of such technology to impersonate someone without consent is illegal and unethical.
For this example, let’s assume we have a fictional AI library named DeepFakeLib that is legally and ethically used for creating deepfakes for consenting individuals or for creating fictional characters.
from DeepFakeLib import DeepFakeModel
# Assume we have a pre-trained deepfake model and consent to use Scarlett Johansson's likeness for demonstration.
# In real-world scenarios, you must have explicit consent or appropriate rights to use someone's likeness.
model = DeepFakeModel(pretrained=True, model_path='/path/to/sc_johansson_model/')
# Load an original video where Scarlett Johansson has given consent to use her likeness.
original_video_path = '/path/to/consented_video.mp4'
# Use the model to generate a new video clip with Scarlett Johansson's likeness.
# Again, this is a hypothetical example; using someone's likeness without their consent is illegal.
fake_video_path = '/path/to/generated_video.mp4'
model.create_deepfake(original_video=original_video_path, output_video=fake_video_path)
# Display the result (in a real program, this would involve a video player component)
print(f"A deepfake video has been created at: {fake_video_path}")This code would theoretically take an input video and then generate a new video that superimposes Scarlett Johansson’s likeness onto a source actor in the original video, resulting in a deepfake.
Note: To reiterate, creating or distributing deepfakes of individuals without their permission is against the law in many jurisdictions and can result in severe legal consequences.
Follow for more things on AI! The Journey — AI By Jasmin Bharadiya
