What Did OpenAI Do This Week? [FREE ISSUE]
An extremely busy week for OpenAI. News count: +20
NEW MODELS, A TIGHTER GRIP, AND A LOT OF LEAKS
OpenAI is pushing out a lot at the moment, with good reason (and some not so good reasons…)
Amidst the sea of plastic vinyl characters being created, OpenAI had an extremely busy week. The company unveiled two new AI models, o3 and o4-mini, that integrate visual reasoning directly into their cognitive processes. These models can manipulate images by cropping, zooming, and rotating as part of their problem-solving approach, enabling a blend of visual and textual reasoning. The o3 model is OpenAI’s most advanced reasoning model to date, capable of handling complex tasks such as coding, mathematics, and visual analysis. Both models are already available to ChatGPT Plus, Pro, and Team users, with o3-pro support expected in the coming weeks. Here are the system cards and the livestream recording. OpenAI also introduced GPT-4.1, a refined version of its flagship model with improved coding capabilities, faster performance, longer context handling, and reduced hallucinations—positioning it as the new default across ChatGPT Pro and API tiers. Here’s the livestream recording of that announcement. OpenAI also introduced Codex CLI, an open-source, terminal-based coding agent that integrates its models directly with local development environments. Designed to execute commands, write and edit code, and even manipulate files from the command line, Codex CLI is a step toward OpenAI’s vision of an “agentic software engineer” capable of handling entire software projects. The tool supports multimodal inputs—such as screenshots or sketches—and is compatible with models like o3 and o4-mini. To encourage adoption, OpenAI is offering $1 million in API grants to eligible development projects. Here’s the OpenAI video walkthrough demo.
Then OpenAI got leaky, and is reportedly developing a social network to compete with X (formerly Twitter), designed around ChatGPT’s image generation capabilities within a social feed. The goal appears to be a platform for sharing and engaging with AI-generated content, creating a rich, closed-loop data ecosystem for further training. While the project has not been officially announced, Sam Altman has been soliciting private feedback, suggesting it’s under active development despite being revealed through leaks.
At the same time, OpenAI introduced new verification rules requiring government-issued ID to access advanced models. The decision follows evidence that rivals like DeepSeek have trained on OpenAI outputs, prompting a clampdown on API abuse. The ID rule represents a shift toward tighter control of access and aligns with broader concerns about safety, misuse, and competitive protection in the AI landscape. Privacy nerds will enjoy arguing both sides of this one…
SO WHAT?
OpenAI is tightening control across its stack. New models (o3, o4-mini) offer solid visual reasoning upgrades but are incremental, not market-shifting. ID verification and VAT enforcement point to a shift from openness to defensiveness—aimed at slowing model mimicry and controlling distribution. Developers and startups relying heavily on OpenAI APIs should reassess risk exposure: costs, access friction, and compliance burdens are rising. Meanwhile, competition from Anthropic, Mistral, DeepSeek, and Groq is accelerating, offering faster, cheaper, and more flexible alternatives (and big tradeoffs).
The leaked social network reveals OpenAI’s real concern: data scarcity. As Reddit, Stack Overflow, and others close off training access, OpenAI is moving to capture user-generated content directly. A social platform gives them proprietary engagement data at scale—but building a sticky, consumer-facing product is a huge cultural and operational leap. If the network is real and works, OpenAI builds in a powerful data moat. If it fails, it wastes resources and alienates its enterprise base. Builders should take note: if you’re not capturing your own data, you’re feeding someone else’s moat.
In the long run, OpenAI wants to own the model, the interface, and the feedback loop. A platform play with, you could say, monopolistic ambitions. But it’s not going to be easy. OpenAI has to watch for overextension, ecosystem backlash, and developer flight. All real risks while the big platforms jostle for positioning. OpenAI, like most AI vendors, are becoming more like cloud platforms: sticky, closed, and expensive. The best advice right now? Stay model-agnostic, build for modularity, prioritise data ownership until you have no choice.
COMPANY ↓
OpenAI, staff/leadership profiles, board & governance, mission & values, HQ & global offices, funding rounds & investors, financial health & revenue, partnerships & alliances, organisational structure & hierarchy, regional presence & expansion
OpenAI CEO Sam Altman revealed that implementing politeness in ChatGPT cost the company tens of millions of dollars. /Evrimagaci
OpenAI identified India as its fastest-growing market, with significant adoption of ChatGPT services. /CNBCTV18
OpenAI entered talks to acquire Windsurf, an AI-powered coding tool, for approximately $3 billion. /Bloomberg
TECHNOLOGY ↓
ChatGPT ecosystem, DALL·E tools, Whisper audio, product roadmap, prototypes, developer access, integrations, extensions, research, innovations
ChatGPT began addressing users by their first names, raising privacy concerns. /AutoGPT
OpenAI launched Flex Processing, offering cheaper and slower AI task processing options. /TechCrunch
OpenAI launched GPT-4.1. /OpenAI
OpenAI released its most advanced AI models yet, O3 and O4 mini, enhancing reasoning and image capabilities. /CNBC
OpenAI discussed the concept of "thinking with images," emphasising the integration of visual understanding in AI models. /OpenAI
OpenAI debuted Codex CLI, an open-source coding tool designed for terminal use. /TechCrunch
OpenAI's new O3 and O4 mini models focused on enhancing image-based reasoning capabilities. /Engadget
ChatGPT introduced an image library feature, allowing users to access and utilize a collection of images. /The Verge
OpenAI announced plans to wind down GPT-4.5, its largest-ever AI model, from its API. /TechCrunch
OpenAI's new GPT-4.1 models focused on coding, offering improved performance and efficiency. /TechCrunch
OpenAI's new GPT-4.1 models emphasised enhanced coding capabilities and long-context comprehension. /Reuters
ChatGPT users are performing reverse location searches from photos, sparking security concerns. /TechCrunch
STRATEGY/IMPACT ↓
Public Policy, CSR, ESG practices, ethics & AI safety, risk management, competitive strategy, thought leadership, industry impact, market positioning & competitive strategy
OpenAI appointed members to its newly formed nonprofit commission to guide philanthropic efforts. /Reuters
Access to future AI models in OpenAI's API may require users to undergo a verified ID process. /TechCrunch
OpenAI stated it might adjust its safety requirements if a rival lab releases high-risk AI models. /TechCrunch
MISCELLANEOUS ↓
Legal, scandals, industrial action, revenue, profiles, public policy, CSR, ESG, corporate culture, ethics, risk management, Sam Altman investments, +
OpenAI's Stargate project announced plans for international expansion, including potential investments in the UK. /TechCrunch + /Reuters
OpenAI announced the enforcement of a 16% VAT on ChatGPT services in applicable regions. /CIO Africa
Microsoft expressed reluctance to continue supporting ChatGPT's training due to resource constraints. /Windows Central
OpenAI began developing a social network to compete with platforms like X (formerly Twitter). /The Verge
The California Attorney General declined to join Elon Musk's lawsuit against OpenAI, citing insufficient public interest justification. /Reuters
A man filed a complaint against OpenAI saying ChatGPT falsely accused him of killing his children. /AOL
RESOURCES ↓
POVs, data visualisations & dashboards, key insights & metrics, industry reports & white papers, competitive analysis, recommended reading & book lists, analysis pieces & case studies +
Alphabet, Nvidia invest in OpenAI co-founder Sutskever's SSI. /Reuters
DeepSeek poses ‘profound’ security threat, US House Panel claims. /Bloomberg
Had this forwarded to you? Want to get it regularly? Smart cookie.
Choose monthly (£15) or annual (£99). Cancel anytime; no monsters here.
> Google more your thing? There’s a WD_DTW? for them too. Go here.
> Amazon more your thing? There’s one for them too. Here it is.