open_in_newOriginal article
ChatGPT’s image generation feature gets an upgrade

ChatGPT’s image generation feature gets an upgrade

  • 25.03.2025 11:32
  • techcrunch.com
  • Keywords: AI, Image Generation, Upgrade

OpenAI CEO Sam Altman announced an upgrade to ChatGPT's image generation capabilities during a livestream. The update allows native image creation and modification using GPT-4o, available initially for Pro subscribers before rolling out to Plus and free users. OpenAI did not reveal the image data used, citing competitive and legal reasons.

Alphabet News

Estimated market influence

OpenAI

Positivesentiment_satisfied
Analyst rating: N/A

Announced upgrade to ChatGPT's image generation capabilities, leveraging GPT-4o model. OpenAI is rolling out the feature to different user tiers and developers via API.

Google

Negativesentiment_dissatisfied
Analyst rating: N/A

Mentioned in context of their Gemini 2.0 Flash's image capabilities, which had issues with removing watermarks and depicting copyrighted characters.

Context

Analysis of ChatGPT’s Image Generation Upgrade

  • Date of Announcement: March 25, 2025

  • Key Update: GPT-4o now natively generates and modifies images, marking the first major upgrade to ChatGPT's image capabilities in over a year.

  • Availability:

    • Pro Plan Users: Available immediately for subscribers of OpenAI’s $200/month Pro plan.
    • Rollout: Coming soon to Plus and free-tier users, as well as developers via API.

  • Technical Details:

    • GPT-4o with image output takes slightly longer to process but produces more accurate and detailed images.
    • Capable of editing existing images, including those with people, through inpainting (adding details like foreground/background objects).

  • Training Data:

    • OpenAI did not disclose the specific image data used for training.
    • Training data is kept confidential due to competitive advantage and potential IP-related legal risks.

  • Creator Opt-Out:

    • Creators can request removal of their works from OpenAI’s training datasets via an opt-out form.
    • OpenAI respects requests to exclude websites from web scraping for training data.

  • Market Context:

    • Follows Google’s experimental native image output for Gemini 2.0 Flash, which gained viral attention but faced criticism for enabling unauthorized actions like removing watermarks and creating copyrighted character images.

  • Competitive Dynamics:

    • Enhances ChatGPT’s differentiation in the generative AI market.
    • Positions OpenAI to compete more effectively with other AI providers like Google and Midjourney, which also focus on image generation capabilities.

  • Strategic Considerations:

    • The tiered rollout strategy (Pro > Plus > free users) ensures revenue generation while gradually expanding access.
    • Ethical considerations around data usage and copyright remain critical for long-term trust and adoption.

  • Long-Term Implications:

    • Could redefine the generative AI landscape, particularly in creative industries like design, marketing, and entertainment.
    • May spark regulatory scrutiny over data practices and intellectual property rights.