AI News, 18 November 2026

Wednesday 18 November

Megaspeed's $12 Billion Nvidia Chip Pipeline for China Exposed as US Tightens Export Rules

AI-generated summary Singapore-based Megaspeed International arranged nearly twelve billion dollars in orders for advanced artificial intelligence chips on behalf of Chinese companies between March 2024 and June 2025, according to records reviewed by The Wall Street Journal. Nvidia stated that it stopped sales and shipments to Megaspeed in June 2025, and a company spokesperson said that Megaspeed remains barred from purchasing the company's chips.

Earlier:

Snowflake Cortex AI Function Studio GA: optimize prompts, compare model cost and quality

AI-generated summary Snowflake made Cortex AI Function Studio and Cortex AI Function Optimization generally available on November 16, 2026. The tools let data teams build custom AI functions in SQL, score them against labeled samples, and optimize prompts or model costs.

Also covered byData Phoenix
Earlier:

The Department of Justice Is Using AI Detector Pangram Against Defense Attorneys

AI-generated summary Department of Justice lawyers are utilizing the AI writing detection service Pangram to argue that opposing attorneys are employing artificial intelligence to write legal arguments. The federal court case involves Khalwinder Bhengura, where government attorneys submitted Pangram results as exhibits to claim the defense generated filings using artificial intelligence.

Reports of AI-generated child sexual abuse material have surged more than 8,400 percent in two years

AI-generated summary Reports of artificial intelligence being used for the sexual exploitation of children increased more than 8,400 percent over a two-year period according to data from Edinburgh's Childlight Institute. The numbers climbed from about 4,700 reported images in 2023 to over 400,000 in 2025 based on reports tech companies filed with the US National Center for Missing & Exploited Children.

Earlier:

European Parliament calls for human oversight of AI in democracy

AI-generated summary The European Parliament adopted a resolution on November 12th, 2026, addressing artificial intelligence and its impact on democratic processes, fundamental rights, and institutional balance. The text calls for human oversight of artificial intelligence within European institutions and proposes a review of the Charter of Fundamental Rights to include algorithmic transparency protections.

Earlier:

AMD completes Taalas acquisition for AI inference

Conceptual illustration of an unbranded dark silicon chip seated above a circuit board with red illuminated traces.

AI-generated summary AMD announced the completion of its acquisition of Taalas to incorporate specialized AI inference silicon and an engineering team. AMD stated plans to integrate the technology into its accelerator roadmap and build system-level products alongside AMD Instinct GPUs.

Also covered byTechPowerUpVenture Heralddailysynapse.com

Nvidia’s $3.5 Billion AI Lease Guarantees

Nvidia’s 2026 chip sales become infrastructure credit support

AI-generated summary Nvidia disclosed $3.5 billion in guarantees for companies leasing land, power, and facilities around its AI chips. The company also offered to rent back unused GPUs from young cloud providers in exchange for a share of their revenue.

Earlier:

Kling Image 4.0: Kuaishou's image model goes native 8K and fits up to 2,000 characters

AI-generated summary Kuaishou unveiled Kling Image 4.0, a new image model featuring native 8K resolution, support for up to 2,000 characters, and a very long prompt limit. Kuaishou says the model is live on the Kling platform and available to businesses and developers through an API.

Also covered bythereelstack.comAITOPAI Ping

Apple Says iCloud+ Plans Will Not Include Increased Siri AI Usage in UK

AI-generated summary Apple announced that iCloud+ plans will not include increased Siri AI usage limits in the UK. Apple stated that daily usage limits will apply to Siri AI and select other Apple Intelligence features starting with iOS 27.3.

Also covered by9to5MacAppleInsiderInc
Earlier:

Detectify launches Agentic Tooling to verify vulnerabilities in live applications

AI-generated summary Detectify has launched Agentic Tooling to let organizations use their own AI agents to investigate and verify suspected vulnerabilities in live applications. Detectify CEO Rickard Carlsson states that the new Security Ledger records testing sessions so teams can verify agent findings against objective evidence.

Earlier:

AI safety legislation has a conflict-of-interest problem — Big US labs are warning of risk just as cheaper Chinese open-weight models erode their commercial edge

Dario Amodei, co-founded Anthropic, looks to Mark Zuckerberg, CEO of Meta Platforms, and U.S. President Donald Trump

AI-generated summary OpenAI launched GPT-6 Astra in September and reported that the model's written reasoning was harder to monitor than its predecessor's. Max von Thun states that commercial considerations loom large for companies like OpenAI and Anthropic as cheaper alternatives emerge.