The Writers Guild of America, East unveiled a partnership to help prevent AI models from training off of its members’ work on Tuesday. The software program Cloak, which creates copies of files that are illegible to AI models but readable for humans, will be available for free to WGAE members as part of the union’s benefits program. The encryption tool will protect writers from their work being exploited by AI without their consent.
“The WGAE partnership with Cloak expands our efforts to protect members’ work and fight back against AI companies exploiting writers and other creative workers,” WGA East president Tom Fontana said in a statement unveiling the partnership. “Members don’t need to worry that their work will be used to train AI models after they send it to producers or agents — or simply that the product of weeks, months, or years of work will be reduced to an AI-generated summary of an email attachment.”
Related Stories
In addition to being provided as a free service to those within the WGAE, writers who aren’t members can add Cloak by registering their script through the guild’s website. The cost to add Cloak protection is $10 for non-members, on top of other pre-existing registration fees.
“For the past three years, the assumption was that anything you put out onto the internet is at risk of being ingested by AI, so I developed Cloak as a way to give agency back to writers to protect against unwanted AI training, AI notes, and analysis,” said Cloak founder Rishi Ambavanekar. “Until now, there hasn’t been a real technically-sound way to fight back against AI and determine how the content you create actually interacts with it. I hope Cloak can keep creative writing in the hands of people and not as data being processed by machines. I want to thank the WGAE and look forward to seeing writers have more confidence in sharing their work far and wide.”
AI’s capacity to dramatically alter and diminish the labor force within the entertainment industry has become an existential concern over recent years. The WGAE and other guilds have sought to include language addressing protections from AI in new union contracts. The Cloak partnership is a product of the guild’s AI Task Force.
“Generative AI models are often falsely advertised on the promise that they will make our work obsolete, but the truth is, the creative process is human and always will be,” said Larry J. Cohen, co-chair of the AI Task Force. “That’s why the guild is advocating for better protections against and awareness of the risks posed by AI — not just through our contracts, but through legislation and member education. Our partnership with Cloak is a natural and necessary extension of our fight to protect our members’ creative rights and the integrity of their work.”
Founded in September 2025, Cloak is backed by the venture capital firm Andreessen Horowitz’s start-up accelerator program Speedrun and the University of Pennsylvania, per the company’s LinkedIn page. The start-up’s website claims that recent tests have proven “effective” against Anthropic’s Claude model up to Fable 5.1, Google’s Gemini model up to 3.8, and OpenAI’s ChatGPT model up to GPT-6 Astra, as well as against automated summaries generated by Google Drive, Gmail, and Adobe Acrobat.