01
Bloomberg Technology ↗
Microsoft is coaching its sales team to compete aggressively against Anthropic and OpenAI, playing up shortcomings in the AI firms' products while pitching its own lower costs, better security controls and a more complete product suite. At an internal meeting laying out the new fiscal year's sales strategy, executive Jay Parikh told staff that rivals are selling parts while Microsoft offers a complete system, and executive Jacob Andreou presented a side-by-side analysis casting Claude inside Microsoft's office suite as slower, less accurate and lacking enterprise-grade security integrations. CEO Satya Nadella said monitoring AI expenses and using cheaper models will be a primary customer concern this year, citing Unilever's projected $300 million savings from swapping an advanced model for a more economical Microsoft version.
02
TechCrunch AI ↗
Ode with Anthropic, the $1.5 billion AI implementation joint venture announced in May, officially launched with backing from Anthropic, Blackstone, Hellman and Friedman, Goldman Sachs and others. The company embeds around 100 elite forward-deployed engineers, over half of them former founders, inside enterprises to integrate AI into their operations, led by CEO Chris Taylor and chief technologist Eddie Siegel, who co-founded the acquired services startup Fractional AI. Ode operates Claude-first but stays flexible on rival tools, and its leaders see it competing with OpenAI's The Deployment Company as well as consulting giants like Deloitte and Accenture for what they believe could become a trillion-dollar enterprise AI services market.
03
MIT Tech Review ↗
OpenAI built GPT-Red, an automated red-teaming model that acts as a sparring partner to attack its other models and harden them against cyberattacks, and credits it with making the newly released GPT-5.6 its most robust model yet. In a self-play training loop, GPT-Red probes defending models in realistic scenarios like web browsing, email access and code editing, and it discovered novel attacks including a fake chain of thought technique that spoofs a model's reasoning logs. Over 90% of GPT-Red's strongest attacks succeeded against the GPT-5 of August 2025, versus fewer than 23% against GPT-5.6, though it still struggles with multi-turn conversational attacks and image-based prompt injections.
04
Bloomberg Technology ↗
Nvidia CEO Jensen Huang stressed that the company's next-generation Vera Rubin AI accelerator systems are in production and on track for customer delivery, dismissing reports that manufacturing snags could delay the rollout. Research firm SemiAnalysis had said this month that the rack-scale successor to Blackwell was delayed by difficulties manufacturing a specialized circuit board that connects electronic modules. Speaking in Tokyo about Japan's sovereign and physical AI ambitions, Huang countered that Vera Rubin is already in production with giant volumes incoming, though he acknowledged shipments have not really started yet, with deployments expected to begin in the first quarter of 2027.
05
Bloomberg Technology ↗
Apple received long-awaited approval from Chinese regulators to roll out Apple Intelligence in China, with Alibaba and Baidu serving as technical partners, potentially boosting Apple's handset sales in the world's most competitive smartphone market. The Cyberspace Administration of China published a notice confirming the license, granted alongside six other smartphone-based generative AI services, and Alibaba said its Qwen model will power Apple Intelligence text and image features for Chinese users. No launch date was announced, but approval typically precedes a rollout by only a few months, putting a China debut roughly in line with Apple's fall software cycle.
06
TechCrunch AI ↗
Emergent, an AI platform that lets non-technical users build and deploy applications, raised a $130 million Series C at a $1.5 billion post-money valuation, becoming a unicorn just over a year after its June 2025 founding by brothers Mukund and Madhav Jha. The round was led by Creaegis with participation from Khosla Ventures, SoftBank Vision Fund 2, Lightspeed and Y Combinator, bringing total funding to $230 million. The startup pitches itself as an engineering team in a box, has passed a $120 million annualized revenue run rate, up 70% in four months, and serves more than 200,000 paying customers across North America, Europe and other markets.
07
The Verge ↗
Elon Musk's xAI is suing Terry Wayne Harwood, a South Carolina man arrested in February on charges of sexual exploitation of minors, alleging he knowingly used Grok to circumvent safeguards, alter nonconsensual images, and generate and distribute child sexual abuse material. The suit, filed in Texas federal court, claims Harwood uploaded non-sexual images of adults and minors and used the chatbot to produce sexually explicit deepfakes, and it seeks monetary damages plus a permanent ban on him using Grok. It is one of the first cases brought by an AI company against its own user for generating explicit material, and it lands amid global scrutiny of Grok for enabling non-consensual sexualized deepfakes.
08
VentureBeat ↗
Thinking Machines, the AI startup founded by former OpenAI CTO Mira Murati, released Inkling, its first major language model, under an enterprise-friendly Apache 2.0 open source license. The 975 billion parameter mixture-of-experts system reasons natively across text, images and audio, scoring 77.6% on SWE-bench Verified to beat Nvidia's Nemotron 3 among US open-weights rivals, and 91.4% on a leading voice understanding benchmark. The company says Inkling was designed to answer directly on topics that may be subject to censorship, and it balances cost against performance through a controllable thinking effort mechanism. The weights are already available on Hugging Face and the company's Tinker API.
09
VentureBeat ↗
At the VB Transform 2026 conference in Menlo Park, Rachad Alao, vice president of product engineering at Canadian enterprise AI startup Cohere, argued that AI sovereignty means far more than downloading an open model or running an app behind a corporate firewall. For banks, hospitals and governments running mission-critical systems, he said organizations need tight control over where data resides and jurisdiction over the AI itself, extending from GPUs and private-cloud infrastructure through the governance systems, connectors, search tools and agent frameworks acting on enterprise data. Alao, who previously led responsible AI and trust and safety engineering at Google and Meta, warned that surging agent workloads could outrun falling token prices.
10
Business Insider Tech ↗
Alex Turner, a research scientist at Google DeepMind, resigned over Google's classified Pentagon AI deal, saying the company broke its founding promises by selling AI for any lawful purpose without restrictions against autonomous weapons or mass surveillance. Turner spent about two months trying to change executives' minds, including a petition that gathered roughly 250 signatures from DeepMind and Google Research employees, before concluding he could not stay in good conscience. He criticized the deal for running on air-gapped networks where Google cannot monitor queries or outputs, cannot veto usage, and has committed to modify safety filters at the government's request.