The AI Brief

Bloomberg reported the talks on August 13, citing people familiar with the matter; Reuters confirmed the figure independently. The deal has not been finalized and talks could fall through, according to people who asked not to be identified discussing private negotiations. If finalized, it would mark Anthropic's largest known acquisition ahead of a hotly anticipated initial public offering.

Decart, founded in 2023 by three Israeli engineers, develops software that improves chip efficiency and reduces AI model training costs, and also possesses world-model generative video technology. The Decart stack spans three lines: DOS, a GPU-optimization layer that speeds up inference; Lucy, a world-editing model that transforms live video for try-on, advertising, and streaming; and Oasis, a world-generation model aimed at robotics and physical AI. DOS 2.0 enables AI agents to process over 1,600 tokens per second, eight times the industry average.

Decart completed a $300 million funding round led by Radical Ventures in May, valuing the company at nearly $4 billion. The proposed $6 billion acquisition price represents a 50% premium to that May valuation. Decart's chip-efficiency software could help Anthropic's existing infrastructure absorb more demand, according to a person familiar with the matter. Upon completion, Decart's team would join Anthropic's inference and performance division. The report comes days after Elon Musk publicly denied that SpaceX was also in talks to acquire Decart.

Disclosure: Claude, which generates this brief, is built by Anthropic.

On August 11, 2026, Google CEO Sundar Pichai announced on X that the Gemini app has crossed one billion monthly active users, making it the fastest-growing product in the company's 28-year history and the 14th Google service to reach that scale. The 1 billion metric applies strictly to the standalone Gemini application on mobile and desktop platforms and excludes users accessing AI features embedded inside other Google products.

Gemini went from 950 million to more than 1 billion in less than a month. To put that pace in perspective, Gmail took roughly 14 years to build an audience that size. ChatGPT reached 1 billion users in June 2026 and still leads overall chatbot share. Google highlighted usage patterns, noting that 63% of Gemini users engage with the assistant through voice, while the platform generates more than 150 million images daily.

What the announcement did not include, how many of those 1 billion users pay for anything, is the question Google chose not to answer; that omission, deliberate or not, is the editorial fact that matters most about the announcement. Google says it now has more than 8 million paid seats across 2,800 companies, with paid enterprise users growing 40% quarter-over-quarter in Q1 2026.

OpenAI's AI ethics lead, Chloé Bakalar, left the company less than a year after joining, according to the Financial Times. Bakalar joined in August 2025 from Meta, where she had been chief ethicist since 2019. A person familiar with her role told the FT that Bakalar was the company's only dedicated ethicist and that OpenAI has no plans to replace her directly. OpenAI did not announce the departure; it became public only when the FT reported it in August.

Her exit follows July's departure of safety systems head Johannes Heidecke, who moved into that role in 2024 when Lilian Weng left. Josh Achiam, who previously led OpenAI's Mission Alignment team before becoming chief futurist, has also reportedly left the company. OpenAI had already disbanded the Mission Alignment team earlier this year. Safety teams have since been folded under Mia Glaese, whose title expanded to VP of Research and Safety.

OpenAI's spokesperson said: "AI ethics doesn't live with one owner or team at OpenAI, and ethical considerations are deeply embedded into the model-building process driven by a number of teams across research." The FT piece does not name a successor or say how the ethics function will be structured next, which is the more useful question for anyone relying on OpenAI's internal review to flag issues before models ship.

Google's Made by Google hardware event in New York on August 12, the most-cited prediction-market window for Gemini 3.5 Pro, passed with no model card, no API entry, and no pricing announcement. The announcement of Gemini's 1 billion users came a day before the event, where deeper Gemini integration across the Pixel 11 series was widely expected. The model remains in limited preview on Vertex AI and has not shipped publicly. First covered in Vol. I, No. 78.

As of August 8, 2026, Gemini 3.5 Pro remains in limited preview on Vertex AI and has not launched publicly. Bloomberg found that Google leaders are worried that Gemini 3.5 Pro may not be competitive with rivals' recent releases. The rebuild decision, reportedly triggered by hallucination rates and real-world reliability issues with the base model, has now produced at least five consecutive missed launch targets since the original June window.

Kalshi prediction market odds place the probability of a Gemini 3.5 Pro launch before August 21 at 63%, and before August 31 at 76%. Google has not confirmed the model's features, pricing, access plans, or rollout schedule. The gap has given Kimi K3, Claude Opus 5, and GPT-5.6 Sol additional runway to sign enterprise customers.

Brad Lightcap, one of the longest-serving executives at OpenAI, announced he is leaving the powerful AI lab to "start something new." Lightcap, who joined the company in 2018, spent four years as OpenAI's chief financial officer. He then ascended to chief operating officer, where he served from 2022 until earlier this year when, amidst a shakeup of executive roles, he transitioned to lead special projects at the company.

Over an 18-month stretch through mid-2025, Lightcap built OpenAI's go-to-market team from around 50 people to more than 700, which included sales reps and staffers focused on customer success, developer relations and strategic partnerships. Denise Dresser, OpenAI's chief revenue officer, has taken over most of Lightcap's responsibilities.

His exit is the latest in a string of departures from OpenAI's senior ranks. Fidji Simo, who had been the company's number-two executive, stepped down in July over health issues. In April, three OpenAI executives announced their departures, including Bill Peebles, who led OpenAI's now-discontinued Sora video app, Kevin Weil, the vice president of OpenAI for Science, and Srinivas Narayanan, CTO of B2B applications. The departures come as OpenAI battles a perception that it's lost ground to rival Anthropic, which is thought to be leading in valuation and enterprise adoption, even as OpenAI's ChatGPT still boasts more users.