1. Anthropic’s $900 Billion Valuation Would Surpass OpenAI for the First Time
Anthropic is raising another $30 billion in a new funding round, with its valuation jumping to $900 billion — surpassing rival OpenAI for the first time. This comes just three months after a round of the same size. Anthropic’s Claude models continue to gain market share in the enterprise sector, driving this valuation leap.
Source: The Decoder (2026-05-16) URL: https://the-decoder.com/anthropics-900-billion-valuation-would-make-it-more-valuable-than-openai-for-the-first-time/
AI Pulse View: Anthropic completing two $30 billion rounds in just three months reflects massive market confidence in its “safety-first” AI approach. Claude’s enterprise penetration is remarkable, and the $900B valuation signals the AI industry is shifting from “OpenAI dominance” to a “multi-player landscape.”
2. OpenAI Launches ChatGPT for Personal Finance with Bank Account Linking
OpenAI has introduced a personal finance feature for ChatGPT, available to Pro users in the US. Through Plaid integration, users can connect their bank accounts to receive personalized financial analysis. ChatGPT can now analyze spending patterns, flag excessive takeout orders, and offer money management advice.
Source: TechCrunch (2026-05-15) URL: https://techcrunch.com/2026/05/15/openai-launches-chatgpt-for-personal-finance-will-let-you-connect-bank-accounts/
AI Pulse View: ChatGPT’s evolution from “chatbot” to “life assistant” is accelerating. Bank account integration means AI will deeply influence personal financial decisions — a leap in convenience that raises serious data privacy and security questions.
3. Greg Brockman Officially Takes Control of OpenAI Products, Unifying ChatGPT and Codex
OpenAI has reorganized its executive ranks once again, with co-founder Greg Brockman now officially overseeing all product operations. The core goal is to unify ChatGPT and Codex into a single product experience, marking the latest in a series of recent restructuring moves.
Source: Wired (2026-05-15) URL: https://www.wired.com/story/openai-reorg-greg-brockman-product/
AI Pulse View: Product unification signals OpenAI’s shift from “parallel product lines” to a “single platform + multiple capabilities” strategy. Brockman’s engineering background could accelerate the architectural integration of ChatGPT and Codex.
4. Musk v. Altman Trial Concludes, Jury Begins Deliberations
The third week of Musk’s lawsuit against OpenAI and Altman has concluded, with both sides trading blows over each other’s credibility. A federal jury is now deliberating. Regardless of the outcome, the trial has profoundly impacted the public image of the AI industry.
Source: MIT Technology Review (2026-05-15) URL: https://www.technologyreview.com/2026/05/15/1137357/musk-v-altman-week-3/
AI Pulse View: This “founder’s dispute” reflects fundamental disagreements about OpenAI’s mission and direction. The verdict will provide important legal precedent for AI industry governance and the path from nonprofit to commercial structures.
5. OpenAI Acquired Voice Cloning Startup Weights.gg, Known for Celebrity Imitations
OpenAI has acquired Weights.gg, a small startup that let users create and share AI voice clones of celebrities including Taylor Swift and Donald Trump. The team of approximately six has now joined OpenAI.
Source: The Decoder (2026-05-16) URL: https://the-decoder.com/openai-bought-a-voice-cloning-startup-famous-for-celebrity-imitations/
AI Pulse View: Voice synthesis is becoming a core competitive battleground for AI giants. Acquiring Weights.gg signals OpenAI’s intent to strengthen its voice technology stack, potentially bringing major upgrades to ChatGPT’s voice interaction capabilities.
6. arXiv Bans Submitters of AI-Generated Hallucinations for One Year
Preprint server arXiv announced it will impose a one-year ban on submitters of AI-generated hallucination papers. A recent surge of low-quality AI-generated submissions has seriously undermined academic integrity and the platform’s reliability.
Source: Ars Technica (2026-05-15) URL: https://arstechnica.com/science/2026/05/preprint-server-arxiv-will-ban-submitters-of-ai-generated-hallucinations/
AI Pulse View: AI is being weaponized to mass-produce academic papers, posing a real threat to the academic ecosystem. arXiv’s ban is necessary but likely insufficient — the academic community needs more systematic AI-generated content detection mechanisms.
7. OpenClaw Founder Runs 100 AI Agents, Spending $1.3 Million Per Month on OpenAI API
Peter Steinberger, founder of the open-source project OpenClaw, runs approximately 100 Codex instances simultaneously with a three-person team for coding, PR review, and bug detection. Monthly OpenAI API spending has reached $1.3 million, which he argues is justified for maintaining innovation velocity.
Source: The Decoder (2026-05-16) URL: https://the-decoder.com/for-1-3-million-a-month-openclaw-founder-peter-steinberger-runs-100-ai-agents-that-code-review-prs-and-find-bugs/
AI Pulse View: This is a landmark case of AI agents at scale. A three-person team orchestrating 100 AI agents represents a new paradigm in human-machine collaboration. The $1.3M/month spend seems staggering but could be cost-effective compared to hiring dozens of engineers.
8. YouTube Opens Deepfake Face-Swap Detection Tool to All Adult Creators
YouTube is opening its Likeness Detection tool to all creators 18 and older. The system identifies AI-generated face swaps in other users’ videos and allows creators to file takedown requests directly.
Source: The Decoder (2026-05-16) URL: https://the-decoder.com/youtube-opens-its-deepfake-face-swap-detection-tool-to-all-adult-creators/
AI Pulse View: YouTube’s move is a direct response to the growing deepfake threat. Democratizing detection tools to everyday creators is an important step toward protecting likeness rights from AI abuse.
9. New AI Security Benchmark: Claude Mythos and GPT-5.5 Can Autonomously Develop Browser Exploits
Researchers at Carnegie Mellon University built a new benchmark measuring how far AI agents can go when exploiting real vulnerabilities in Google’s V8 engine. Claude Mythos leads GPT-5.5, showing significant progress in autonomous cybersecurity capabilities.
Source: The Decoder (2026-05-16) URL: https://the-decoder.com/new-benchmark-shows-claude-mythos-and-gpt-5-5-can-develop-real-browser-exploits-autonomously/
AI Pulse View: AI’s ability to autonomously develop exploits is both a powerful security research tool and a potential threat. As AI agent capabilities advance, balancing “using AI to find vulnerabilities” against “preventing AI from being weaponized” will be a critical challenge.
Other Developments
- MIT Technology Review covered how Chinese short dramas have become AI content factories, showcasing new applications of AI in the entertainment industry (2026-05-15)
- Wired reported on Mira Murati (Thinking Machines Lab founder, former OpenAI CTO) stating she wants AI to “keep humans in the loop” rather than automate people out of jobs (2026-05-15)
- Ars Technica featured an interview with Claude Code’s product lead discussing usage limits, transparency, and the “lean harness” (2026-05-15)