Assemblyman Marc Berman (D-San Francisco) has introduced a bill, AB 602, which would create criminal liability for the creation and distribution of “deceptive recordings,” commonly known as “deepfakes,” if the recording is likely to deceive any person who views the recording, or to defame, slander, or embarrass the subject of the recording.
The bill defines as deceptive recording as “any audio or video recording that has been created or altered in such a manner that it falsely appears, to a reasonable observer, to be an authentic recording of a person’s actual speech or actions.” The penalty for violation of the bill’s provisions is up to a year in jail and/or a $2,000 fine.
AB 602 comes about in response to new technologies that have made it possible to create hyper-realistic video and audio recordings that appear to show individuals doing or saying something that they did not do or say. As reported by the Washington Post in a recent expose, this technology has been used to create sexually explicit materials using the co-opted images of women found on the internet. Concerns have also been raised that the technology could be used to incite violence and interfere in elections by making it easier to pass off fake events as real.
While the potential for abuse of “deepfake” technology is certainly concerning, the First Amendment implications of attempts to regulate use of the technology is likewise troubling. CNPA has expressed serious concern about the constitutionality of the bill to Assemblyman Berman’s office, and has also urged the author to consider the chilling effect the bill would have on speech even if it was amended to ameliorate constitutional concerns.
Because of these constitutional concerns, and the fact that existing false light, defamation, and copyright laws already prohibit many of the problematic uses of “deepfake” technology, CNPA plans to oppose AB 602 as it is currently written. CNPA will continue to work with Assemblyman Berman to determine whether the bill can address the problematic uses of “deepfake” technology without harming First Amendment protected activities.