🔥 The Week in AI News
- 4 days ago
- 3 min read
OpenAI, Adobe, Google, Suno, ChatGPT, Apple, MiniMax, Meta, Shopify, Anthropic, Spotify, Merlin, Reddit, ChatGPT, GPT-5.6 Luna, Photoshop, Acrobat, Adobe Express, Gmail, Google Calendar & more.
▪️ OpenAI gave free users unlimited ChatGPT text conversations.
OpenAI removed the text-chat ceiling for Free and Go users, with GPT-5.6 Luna becoming their new default model and a Think button available for more complex questions. Files, images, voice and image generation still have limits, but everyday ChatGPT use just became dramatically less restricted.
▪️ Adobe brought Photoshop, Acrobat and Express directly into ChatGPT.
People could edit images, modify PDFs and create social graphics by describing what they wanted without leaving the conversation. This wasn’t the full Adobe desktop experience, but it pushed ChatGPT closer to becoming the place where creative work actually gets completed—not just discussed.
Ask Maps gained the ability to order food, compare hotels, find tickets, monitor transit and personalize recommendations using information from Gmail and Google Calendar. Google Maps wasn’t just showing people where to go anymore, it was starting to handle what happened when they arrived.
▪️ Suno moved to watermark AI music and restrict how generated songs could spread.
The company announced new watermarking, fingerprinting and transparency systems designed to identify Suno-generated music and combat spam, fraud and misuse. It also planned tighter download rules, because unlimited synthetic music flooding every platform was never going to remain a charming little experiment.
ChatGPT Work shifted the model from answering isolated questions toward gathering context, creating finished deliverables and managing recurring workflows. The real opportunity wasn’t writing increasingly elaborate prompts—it was giving ChatGPT a defined outcome, clear standards and an actual piece of work to finish.
▪️ Gen Z dating apps replaced endless swiping with AI matchmaking.
Apps like Ditto used conversational onboarding to understand users, select matches and arrange real-world dates at specific times and locations. After years of turning romance into an infinite card-sorting exercise, dating apps finally started experimenting with getting people off the app.
▪️ OpenAI argued that Apple’s own security failures weakened its trade-secrets lawsuit.
OpenAI claimed Apple allowed employees to use personal iCloud accounts for work and failed to properly revoke access when they left the company. The defense was essentially that Apple couldn’t treat information casually internally and then suddenly describe it as Fort Knox-level intellectual property in court.
Hailuo MiniMax 3 combined text, image, video and audio references to create up to 15 seconds of 2K video with native stereo sound. The strongest results came from giving every input one clear job, because stuffing an entire blockbuster into a single generation remained the fastest way to turn credits into expensive nonsense.
▪️ Meta launched Muse Code to handle engineering work across massive codebases.
The terminal-based agent could plan changes, write code and validate its work while launching parallel sub-agents inside isolated worktrees. Meta wasn’t trying to build another autocomplete tool—it was going after the larger software-engineering workflows already being targeted by Codex and Claude Code.
▪️ Shopify said AI search tripled traffic and orders without replacing Google.
AI-driven visits and purchases to Shopify stores reportedly tripled year over year while traditional search traffic continued growing. AI assistants were especially useful for smaller merchants because they could match products to detailed intent instead of simply rewarding whatever already ranked highest for a broad keyword.
▪️ Anthropic started assembling a team to design custom AI chips.
The company began hiring engineers to co-design hardware and models that could make Claude faster, more efficient and less dependent on outside chip suppliers. The AI race had already moved beyond model quality—the companies that controlled their own compute increasingly controlled their own future.
▪️ Spotify brought independent labels into its licensed AI remix experiment.
A deal with Merlin opened Spotify’s upcoming fan-made covers and remixing tools to participating artists from more than 30,000 independent labels. The system was designed around consent, attribution and compensation, offering a much smarter path than flooding streaming platforms with anonymous synthetic artists and calling it innovation.
New EU AI Act transparency rules required companies to disclose AI interactions and identify realistic content that had been generated or manipulated. Machine-readable markers also became part of the equation, turning provenance and disclosure from optional ethics language into actual product requirements.
Brands and marketing agencies increasingly targeted Reddit with AI-generated posts and stealth promotions designed to influence answers from ChatGPT, Gemini and other AI search tools. The next SEO war wasn’t only about ranking on Google—it was about manufacturing the supposedly authentic human conversations that AI systems trusted.
Comments