DX Today AI Daily Brief - Sunday, May 10, 2026

DX Today AI Daily Brief - Sunday, May 10, 2026

Anthropic edges toward a one trillion dollar valuation as it pursues a fifty billion dollar capital raise, while also locking down a one point eight billion dollar compute deal with SpaceX Colossus One. Sony and TSMC unveil a strategic AI image sensor partnership for autonomous systems. Apple prepares to let users select their preferred AI model in iOS 27, opening Siri to Gemini, ChatGPT and Claude. Snap and Perplexity quietly end their four hundred million dollar deal. Microsoft acknowledges its 2030 carbon negative pledge is under significant strain from AI energy demand. Sierra raises nine hundred fifty million dollars at a fifteen billion dollar valuation, while China-based Moonshot AI closes two billion dollars at a twenty billion dollar valuation. Meta expands AI powered teen safety protections globally. Americans push back on hyperscale AI data centers as startups pitch tiny in-home alternatives. Nvidia tops forty billion dollars in 2026 AI equity bets. And Brussels strikes a provisional deal to streamline the EU AI Act.

DX Today - Your daily AI news briefing, delivered every morning.

Send us fan mail

#AINews #ArtificialIntelligence #Anthropic #Nvidia #DXToday

Transcript

SPEAKER_04 0:00

It's Sunday, May 10, 2026. You're listening to the DX Today AI Daily Brief. Today, Anthropic edges toward a$1 trillion valuation. Apple prepares to let users pick their preferred AI model inside iOS 27. And a Wall Street giant retrenches as AI spending strains a long-standing climate pledge. Let's get into it.

SPEAKER_00 0:27

Anthropic is in talks for a fresh capital infusion that could push the maker of Claude to a valuation approaching$1 trillion, according to a Financial Times report on May 8th. The proposed round would bring in roughly$50 billion in new capital at a pre-money valuation in the$850 to$900 billion range. That would represent more than a doubling of the company's$380 billion mark from its February Series G. Sources say no term sheet has been signed, and discussions remain ongoing. If completed, the round would lift Anthropic past opendate AI in private market value, an extraordinary reversal of fortune in just three months.

SPEAKER_04 1:09

From valuations to compute.

SPEAKER_02 1:11

Anthropic also moved aggressively on infrastructure this week, signing a roughly$1.8 billion deal with SpaceX's Colossus One Data Center. The agreement gives Anthropic over 300 megawatts of additional compute, which the company says is equivalent to more than 220,000 NVIDIA GPUs coming online in the next month. Tucked at the bottom of the announcement, Anthropic said future collaboration may include orbital AI compute capacity measured in multiple gigawatts. The deal also doubles clawed code rate limits for developers. It marks one of the largest single compute purchases in AI history and intensifies a growing race for power that is reshaping global infrastructure planning.

SPEAKER_03 2:03

Now a milestone in chip making. Sony and Taiwan Semiconductor Manufacturing Company unveiled a strategic partnership on May 9th, focused on next-generation AI image sensors for autonomous systems. The collaboration pairs Sony's leading sensor design expertise with TSMC's advanced manufacturing nodes to deliver chips purpose-built for autonomous vehicles, robotics, and industrial vision. Sony shares spiked on the news, with trending interest jumping more than 100%. The companies say the partnership is intended to accelerate AI perception capabilities across rapidly expanding markets where image processing latency is now a critical performance bottleneck. Industry analysts describe the deal as a turning point in the consolidation of compute and sensors for embodied AI. And Apple opens its doors.

SPEAKER_01 2:57

Apple is planning to overhaul Siri and let users select their preferred AI model inside iOS 27, according to FreshLeaks reported May 9th. Customers will be able to choose Google Gemini, Open.ai's ChatGPT, or Anthropics Claude, as the default supporting AI service for features such as writing tools, image playground, and search. A new framework called Extensions will allow developers to plug in generative AI capabilities through Apple Intelligence on Demand. The shift, expected to be unveiled at WWDC next month, marks Apple's most significant strategic concession yet. The company is signaling that openness, not exclusivity, is the path forward for its AI platform.

SPEAKER_04 3:45

A partnership unravels.

SPEAKER_05 3:47

Snap confirmed this week that its previously announced$400 million partnership with Perplexity has ended before any broad rollout. The deal, originally hailed as a marquee distribution agreement that would bring Perplexity's AI answers natively into Snapchat, only progressed through limited testing. Snap disclosed the development on its first quarter earnings call. Both companies have declined to share specifics about why the agreement collapsed, but analysts point to user adoption challenges and unresolved revenue share questions. The dissolution is a notable setback for perplexity as it competes with OpenAI and Google for distribution. It also raises questions about the durability of high-profile AI tie-ups.

SPEAKER_04 4:33

Climate goals feel the strain.

SPEAKER_00 4:36

Microsoft acknowledged on May 8th that its commitment to be carbon negative by 2030 is now under significant pressure due to the energy-intensive buildup of AI infrastructure. The company's 2025 Environmental Sustainability Report shows total carbon emissions are up nearly 24% since 2020, driven primarily by scope 3 emissions from data center construction. CEO Satya Nadella has pledged to double AI infrastructure capacity within two years, even as power demand from training and inference balloons. The report acknowledges the climate costs while reaffirming the company's 2030 target as aspirational. Microsoft says new nuclear, geothermal, and grid scale storage agreements will help close the gap.

SPEAKER_04 5:21

A mega round for customer experience.

SPEAKER_02 5:23

Sierra, the AI customer experience startup, co-founded by former Salesforce co-CEO Brett Taylor, raised$950 million in fresh funding on a$15 billion valuation. Google Ventures and Tiger Global led the round. Sierra has deployed its AI agents inside hundreds of enterprise customer service operations, automating tasks ranging from billing inquiries to technical support. The new capital arrives just weeks after Sierra reported tripling year-over-year revenue and signing flagship contracts with major retailers and airlines. The fundraise represents a near doubling of Sierra's valuation in less than nine months. It cements the company as one of the most valuable AI-native startups in the customer experience category.

SPEAKER_03 6:20

Now eastward to Beijing. Chinese AI Lab Moonshot, the maker of the Kimi Chatbot, closed a roughly$2 billion funding round at a$20 billion valuation. The deal was led by Maitwan's venture arm, Longzi Investments, with participation from Tsinghua Capital, China Mobile, and CPE Yuanfeng. Moonshot's OpenWaite, Kimi K2.5 and K2.6 models have surged to become some of the most heavily used large language models on developer platform OpenRotter. The company reported annualized revenue exceeding$200 million in April. Moonshot has now raised more than$3.9 billion over six months, reflecting fast accelerating Chinese investor appetite for sovereign open source AI champions.

SPEAKER_04 7:12

Protecting young users online.

SPEAKER_01 7:14

Meta this week announced an expansion of its AI-powered teen protection program. Reports surfaced May 9th that the company is now using machine learning models to identify users between 13 and 17 and apply stricter privacy and safety defaults through teen accounts. The system uses contextual profile analysis and visual age estimation to flag suspected age misrepresentation. Meta has rolled the protections out across the European Union, Brazil, and Facebook in the United States. The announcement comes amid intensifying regulatory pressure on social platforms over youth safety. Meta says these AI-assisted controls will become the default for new Instagram and Facebook accounts, with appeals available for users wrongly identified as miners.

SPEAKER_05 8:03

On the home front, a CNBC report on May 9th highlights mounting public opposition to large AI data center construction across the United States. Voters in towns from Virginia to Arizona have pushed back on hyperscale projects, citing concerns about water use, noise, and grid reliability. But a small wave of startups is pitching an alternative. Tiny in-home AI compute units the size of toasters that run inference workloads on residential power. The pitch resonates with consumers who oppose corporate megacenters but welcome decentralized hardware paying them for spare cycles. Maine recently saw its governor veto a hyperscaler ban. The story illustrates a deepening political flashpoint over where AI compute should live.

SPEAKER_04 8:54

Nvidia goes shopping.

SPEAKER_00 8:56

A CNBC analysis on May 9th showed that NVIDIA has now committed more than$40 billion to AI equity investments across 2026. This week alone, NVIDIA secured the right to invest$2.1 billion in data center operator IRENN to deploy up to 5 gigawatts of DEISX aligned AI infrastructure. A day earlier, the company announced an option to invest$3.2 billion in glassmaker Corning to scale specialty optics for AI networking. Other recent bets include$30 billion in Open Data AI,$2 billion in Nebius, and stakes in Anthropic and XAI. Analysts say Nvidia is using its market cap to ensure long-term demand for its chips.

SPEAKER_02 9:40

And from Brussels, European Union negotiators on May 7th reached a provisional agreement to simplify the Blocks Landmark AI Act in a streamlining package known as Omnibus 7. The deal extends regulatory exemptions for small and mid-cap companies, reduces requirements in narrow cases, expands powers for the new AI office, and reinforces transparency obligations. It also adds a new prohibition targeting non-consensual intimate content and child sexual abuse material generated by AI. The deadline for national AI regulatory sandboxes has been pushed to August 2, 2027. The agreement still requires formal sign off from member states. It marks the most significant adjustment to European AI law since the Act first entered force last year.