"It's a blatant abuse of the legacy of one of the greatest hip-hop artists of all time. The estate would never have given its approval": A 'flagrant violation' and AI feature Tupac Shakur in a "Taylor Made Freestyle" track 28 years after his demise

What you need to know

• Canadian rapper Drake is in trouble for releasing a diss track featuring an AI-generated verse from deceased hip-hop star Tupac Shakur aimed at Kendrick Lamar.
• Tupac's estate has sent a cease and desist letter demanding the track to be pulled down or it'll move to court for legal action against the flagrant violation. 
• The track has sparked a debate across social media platforms over its authenticity, flow, and more. 
• Drake has since pulled down the track with AI-generated verses from Tupac Shakur and Snoop Dogg.

The emergence of generative AI has unlocked a new realm full of untapped opportunities and potential. Who'd have guessed that average users could generate Hollywood-like short clips from the comfort of their homes using sophisticated software like OpenAI's Sora? 

On the flip side, the lack of elaborate measures, regulations, and guardrails presents a grey area that doesn't clearly define what AI-powered tools can be used for. This is one of the main reasons why there's a sudden uptick in deepfake propaganda about the impending US Presidential election. You might also remember pop star Taylor Swift's AI-generated fake images that surfaced across social media platforms, sexually depicting her.

More recently, Canadian rapper and singer Aubrey Drake Graham (better known as Drake) landed himself in trouble for releasing a track that featured an AI-generated verse from Hip-hop star Tupac Amaru Shakur who was tragically killed in 1996.

According to Billboard, the late Tupac Shakur's estate recently sent a cease and desist letter to the Canadian rapper demanding the removal of the diss track (Taylor Made Freestyle) addressed to Kendrick Lamar within 24 hours or it will move to court for legal action.

Howard King, the lawyer representing Tupac's estate indicated:

"The estate is deeply dismayed and disappointed by your unauthorized use of Tupac’s voice and personality. Not only is the record a flagrant violation of … the estate’s legal rights, it is also a blatant abuse of the legacy of one of the greatest hip-hop artists of all time. The estate would never have given its approval for this use."

Even as I went to YouTube to listen to the track, the ad outlined how to generate AI music for YouTube videos before the music played. While YouTube's ad issue is a major concern, that's a story for another day. 

I'd have never guessed Tupac's verse was AI-generated

Aside from following technological advances, I double up as a hip-hop head and probably spend too much time listening to music than I should. I'm also a huge fan of Tupac as I grew up listening to his tracks and do so to date. 

I just learned about this bizarre story a few hours ago. And frankly, if someone sent me this track to listen to before finding out it was AI-generated, I wouldn't have guessed it in a million years. This depicts how powerful AI is and how hard it is to tell what's true and fake.

However, after giving it a couple of listens I gradually picked up several things that were off. Any hip-hop lover and fan of Tupac's work will be able to tell the flow of the verses, transitions, and more is different. As reiterated by a user who also flagged this issue in the video's comment section on YouTube "Stealing Tupac's voice and still sounding like Drake is wild."

In a similar predicament, the diss track spurred different reactions from hip-hop lovers worldwide. Another user indicated "From having ghostwriters to write bars to having bars written for ghosts, Drake came a long way."

Tupac's lawyer added:

"The unauthorized, equally dismaying use of Tupac’s voice against Kendrick Lamar, a good friend to the estate who has given nothing but respect to Tupac and his legacy publicly and privately, compounds the insult.”

Companies deeply invested in AI-powered advances like OpenAI have highlighted their efforts potentially designed to resolve this issue. In February, the ChatGPT maker announced the commencement of watermarking images generated using ChatGPT or API serving the DALL-E 3 technology. However, it admitted that the step was "not a silver bullet to address issues of provenance."

Drake has already pulled down the track.