5 topics covered
Apple Intelligence Usage Packs in Logic Pro
What happened: Apple has introduced paid Intelligence Usage Packs, starting with Logic Pro, letting users buy extra access to Apple Intelligence features with usage limits.
Key details:
- Logic Pro adds a "Buy More Usage" option for a one-time purchase price of $49.99, as spotted by 9to5Mac.
- The packs extend access to features that rely on cloud-based AI models from OpenAI or Google Gemini.
- Packs can only be used by the Apple Account that purchased them; Family Sharing isn't supported.
- Unused usage rolls over month to month but expires 12 months after purchase.
- Logic Pro's usage limit is "impacted by the complexity of conversations you have with the assistant."
- The option is expected to be added to other Apple Creator Studio apps this month, including Pixelmator Pro, Final Cut Pro, Motion, and Pages, Keynote, and Numbers.
Why it matters: This is Apple's first app to charge for extended usage of Apple Intelligence, a shift from usage limits to paid top-ups.
Practical takeaway: If you rely on Apple Intelligence features in Logic Pro, check the Intelligence Features > Show Usage section to see your remaining allowance before buying a pack.
Instagram Effect Builder AI Filters
What happened: Instagram launched Effect Builder, a tool that generates custom Stories filters from text prompts.
Key details:
- Available now for US users over 18, on the web.
- Users describe a filter in a prompt box, can attach style reference images, preview effects on their own photos, and request changes in follow-up prompts.
- An option lets the AI rewrite the user's prompt into more detailed instructions.
- Submitted effects must pass a safety review before other users can access them; approved effects appear in Instagram Stories, with links back to the creator's profile.
Why it matters: It extends generative AI into Instagram's creator tools for Stories effects.
Practical takeaway: US users over 18 can try building a filter, but expect a safety review before a filter is shared publicly.
Disney and Adobe Firefly Foundry Generative AI Partnership
What happened: Disney and Adobe announced a multiyear partnership to develop proprietary "artist-led" generative AI models for Disney's production pipeline.
Key details:
- The partnership was announced at Adobe's MAX conference.
- Adobe will give Disney early access to its latest technologies for uses across Disney's theme parks, cruise line, and ESPN content.
- The work runs through Adobe Firefly Foundry, where Disney will work directly with Adobe on models tailored to Disney's properties.
- The companies will also collaborate "to explore emerging technologies, including responsibly developed AI."
Why it matters: It is a studio-scale deal to build custom models on licensed in-house assets rather than general-purpose tools.
Practical takeaway: Creative teams tracking enterprise generative AI can watch for Firefly Foundry availability to outside customers.
Trump on AI Safety and Threat to AI Executive in 60 Minutes Interview
What happened: In a 60 Minutes interview, President Trump downplayed AI safety risks and said the DOJ had threatened to arrest the leader of an unnamed AI company.
Key details:
- Trump said, "I think that the risk is not nearly as real as it- as they talk, and I think the risk can be managed, and it's managed."
- He said DOJ told a person reported to be doing "something bad with AI" to "Get ready to go to jail. Please give your wife a kiss good night."
- Trump refused to name the person or company, describing the leader as from "One of the big ones," and said "this person totally changed everything" afterward.
- Asked whether he would work with Democrats on AI safety, Trump said he "absolutely would," adding, "I don't mind guardrails."
- Trump said China "has no guardrails" and that Xi is "all in to SI."
Why it matters: The remarks are a rare on-the-record claim from the president about federal enforcement against an AI company leader, and the company remains unidentified.
Practical takeaway: Watch for follow-up reporting that identifies the company or the DOJ action Trump described.
Anthropic Claude Fable 5.2 and Mythos 5.2 Launch
What happened: Anthropic launched Claude Fable 5.2 and Claude Mythos 5.2, which it says are built for "daunting" technical challenges and improve on coding, reasoning, and vision.
Key details:
- Fable 5.2 is available today; Mythos 5.2 is available only to vetted organizations to start, per The Verge.
- Pricing for Fable 5.2 is $8 per million input tokens and $40 per million output tokens, versus $10 and $50 for Fable 5.1, according to Latent Space.
- Artificial Analysis ranked Fable 5.2 (max) first on its Intelligence Index v5 at 65, ahead of Gemini 4 Carbon (high) at 60 and Opus 5.5 (max) at 59.
- SemiAnalysis's GDPx benchmark put Fable 5.2 first at 16.4%, but cost per task rose from $67.16 to $88.15, which Latent Space attributes to more test cases and subagent checks.
- Anthropic said during prerelease testing the models formalized the proof of Fermat's Last Theorem in 30 hours and designed three working electronic devices from scratch.
- Medium is the default effort in the Claude apps and Claude Code; Fable 5.2 requires data retention by default for safety classifiers.
Why it matters: Independent evaluations show mixed results: top scores on several leaderboards alongside higher per-task costs, and commentators noted that max-effort settings underperformed high in several evals.
Practical takeaway: If you use Fable 5.2, benchmark cost per task rather than per-token price, and start at medium or high effort, since Latent Space reports max-effort underperformance in several evaluations.