As artificial intelligence companies face a glut of lawsuits alleging copyright violations, some music industry veterans are wondering if the distinctive nature of their voice might bolster any claims in court.

The idea, as reported in The Verge, is that rather than using copyright, some in the industry have begun suggesting that an artist can use their likeness to protect their work against deep fakes. There’s legal precedent: Bette Midler successfully sued Ford after the car company used a sound-alike singing a song from one of her albums in an ad. Tom Waits won a similar suit against Frito Lay.

"Likeness is definitely the first place you would look to when forming new legislation because an artist’s likeness is based on their voice," Evan Dhillon, founder of voice cloning platform Kits.ai told The Verge.

The issue has risen to the fore as artificial intelligence tools that clone or replicate an person's voice are proliferating. While some use the technology for comedy (witness the Beach Boys singing Nine Inch Nails’ 'Hurt'), others have used it create new material. In May, an AI-generated song was uploaded to Spotify and YouTube called Heart on My Sleeve that featured a virtual Drake and The Weeknd. The song was later removed because the creator hadn’t licensed a sample used in the music — not because it used Drake and The Weeknd's voices.

Music copyright cases can be difficult to adjudicate. In western music, there are only 12 notes; some commonalities are inevitable. For instance, it’s impossible to copyright a chord or even a chord progression (myriad songs use the exact same chords in the exact same order).

Other legal gray areas remain: for instance, a record company can hold the rights to an artists’ songs, but licensing their voice is a fundamentally different issue.