Meta's New AI Tool Creates Deepfakes. Here's How to Protect Yourself on Instagram

Summary

Meta's new AI tool, Muse Image, can create deepfakes using users' likenesses. Users can opt out by making their accounts private or disabling content reuse for AI features in Instagram settings. Advocacy groups have voiced strong concerns about potential exploitation and misuse.

Key Takeaways

1. Meta's new AI model, Muse Image, allows users to create hyperrealistic AI images, including deepfakes, of others by including their username in prompts.
2. Users with public Instagram accounts can opt out of being included in AI creations by either making their account private or disabling the 'Allow people to reuse your content on Instagram and with AI features at Meta' setting.
3. Advocacy groups like SAG-AFTRA and the National Center on Sexual Exploitation have expressed strong criticism, citing potential for exploitation, abuse, harassment, and identity fraud.

Statement Breakdown

• Claimed Facts: 50% of statements the article presents as facts
• Opinions: 30% of statements classified as editorial or subjective
• Claims: 20% of statements surfaced for additional reader evaluation

Credibility & Bias Reasoning

Credibility assessment: The article presents factual information about Meta's new AI tool and its potential implications. It includes direct quotes from relevant organizations and provides actionable advice for users. However, it relies on some speculative statements about future misuse and the effectiveness of safeguards.

Bias assessment: Concerned Tech Critic. The article adopts a critical stance towards Meta's new AI tool, highlighting potential harms and risks. It emphasizes negative reactions from advocacy groups and expresses skepticism about the effectiveness of Meta's safeguards, framing the tool as a 'deepfake disaster.'

Note: This article raises valid concerns about AI deepfakes and user privacy. While it provides practical steps for protection, some claims about potential misuse are speculative. Readers should consider the author's critical perspective.

Credibility flag: Caution Advised

Claimed Facts (10)

• This is a factual statement about the origin and naming of the AI model.
• This states the current and planned availability of the AI tool.
• This describes the mechanism by which deepfakes can be created using the tool.
• This is a factual instruction on how to change account settings.
• This is a factual instruction for navigating the Instagram app.
• This is a factual instruction for navigating the Instagram app.
• This is a factual instruction for navigating the Instagram app.
• This is a factual instruction for changing account settings.
• This is a factual instruction for navigating the Instagram app.
• This is a factual instruction for changing account settings.

Opinions (5)

• This statement expresses a strong negative opinion and uses loaded language ('disaster', 'slop').
• This is a strong recommendation framed as a necessity, implying a high degree of risk.
• This is an assertion about the importance of the controls for a specific group, framed as a certainty.
• This expresses skepticism and uncertainty about the effectiveness of Meta's safeguards.
• This is a statement about a consequence of making an account private, presented as a definite negative outcome.

Claims (5)

• While plausible, this is an anecdotal claim presented as evidence of ease of misuse without independent verification.
• This quote contains strong, emotionally charged language ('utter miscalculation', 'obvious dangers') that may be intended to provoke a strong reaction rather than present a purely factual assessment.
• This quote uses rhetorical questions and strong accusations ('obvious and foreseeable opportunities') that lean towards alarmism and may not be fully substantiated.
• This is a generalization based on a past event (Sora) and asserts that bad actors 'can' bypass safeguards, which is a strong claim about future or potential actions.
• This is presented as a claim by Meta, but the article immediately casts doubt on its effectiveness, making it a point of contention rather than a settled fact.

Key Sources

• Katelyn Chedraoui — Author, CNET
• SAG-AFTRA — Actors union
• LA Times — Newspaper
• Haley McNamara — Executive Director, National Center on Sexual Exploitation
• National Center on Sexual Exploitation — Tech watchdog
• OpenAI — AI Company
• Meta — Technology Company