AI News Today - June 10, 2026: 16 Biggest Stories

Tuesday, June 10, 2026. This is one of the most consequential 48-hour windows in AI history. Yesterday, Anthropic launched Claude Fable 5 - the first Mythos-class model available to the general public, setting a new benchmark record on SWE-Bench Pro with 80.3%. Today, SpaceX is one day from its $135/share IPO pricing that will determine the valuation benchmark for the entire AI IPO wave. Apple is locked in an escalating standoff with the EU over the Siri AI rollout, raising questions about the $1 billion/year Gemini deal's European reach. Taiwan is considering sweeping AI chip export restrictions on China. A $200 million AI robotics startup just emerged. And the brain-inspired AI sector just absorbed a $500 million bet from Jeff Bezos. Here are all 16 stories, every one sourced, ranked by signal strength.
1. Claude Fable 5 - Anthropic Releases First Mythos-Class AI for General Use

On June 9, 2026, Anthropic launched Claude Fable 5 - the first model from its Mythos tier to be made available for general use. The official announcement on Anthropic's website describes Fable 5 as sitting above the Opus class in capability, with a name drawn from the Latin fabula ("that which is told"), akin to the Greek mythos. Until yesterday, every Mythos-class model had been locked behind Project Glasswing's tightly controlled access program. Fable 5 is the same underlying model as Claude Mythos 5, the difference being safety classifiers that redirect cybersecurity and biology queries to Claude Opus 4.8 rather than answering directly.

Anthropic's framing of the release is the most ambitious it has ever used for a public launch: Fable 5 is described as "state of the art on nearly every benchmark we tested." The company says this model is capable enough that, on certain topics, queries will be silently rerouted to a different model - a signal that the capabilities available here, without the safety classifiers, are considered too powerful for unrestricted deployment.

Stripe conducted early testing and reported substantial improvements in software engineering workflows. Matthew Pines, testing frontier physics research, reported Fable 5 reached in 36 hours what GPT-5.5 achieved in four days on the same task. The pattern across independent testers is consistent: the model stays on task longer, validates its own work before declaring it done, and holds focus across longer task horizons. PCWorld's hands-on report describes it as "Mythos-class... with the potential to do profound good for the world."

2. Claude Fable 5 Benchmarks - SWE-Bench Pro 80.3%, GDPval-AA 1932, Terminal-Bench 88.0%

The benchmark results from Anthropic's Fable 5 / Mythos 5 comparison table (June 9, 2026) show a model that leads the field on agentic coding by a substantial margin. On SWE-Bench Pro - the industry-standard agentic software engineering benchmark - Claude Fable 5 scores 80.3% compared to GPT-5.5's 58.6%, a gap of over 21 percentage points. On Terminal-Bench 2.1 (Mythos 5, with safeguards lifted in relevant areas), the score reaches 88.0%. GDPval-AA - a knowledge-work evaluation - reaches 1932, up from Claude Opus 4.8's 1890.

Additional benchmarks from the Vellum.ai benchmark breakdown (June 9, 2026): spatial reasoning improved significantly; tool use score up 17.4% versus Opus 4.8; legal benchmark up 13.3%; health benchmark at 66.0% (Mythos 5). FrontierCode Diamond: 29.3%. OSWorld-Verified: 83.4%. Fable 5 is built for tasks where complexity and duration are the primary variables - large-scale code migrations, multi-day agentic sessions, deep research, and dense knowledge work. Anthropic's stated principle: "the longer and more complex the task, the larger Fable 5's lead over our other models."

The asterisked scores - Terminal-Bench 2.1 (88.0%) and Health (66.0%) - are Mythos 5 scores. Fable 5 lands closer to Opus 4.8 on those metrics because the safety classifiers redirect those query types before Fable 5 processes them. This means the benchmarks slightly understate the raw model capability of Fable 5 in areas where safety classifiers activate.

For a model selection guide - when to use Fable 5 vs Opus 4.8 vs Sonnet 4.6 by task type, token budget, and latency requirements - the Claude AI Complete Hub covers the full Anthropic model lineup with use-case routing guidance.

3. Claude Fable 5 Pricing, Access Window, and Platform Rollout

Pricing is $10 per million input tokens and $50 per million output tokens -- double the price of Claude Opus 4.8 ($5/$25). This puts Fable 5 above every other publicly available model on a per-token basis. Prompt caching delivers a 90% discount on cached input tokens, which partially mitigates costs for applications with large, reused system prompts. Per Finout.io's pricing analysis, the benchmark evidence -particularly SWE-Bench Pro (80.3% vs GPT-5.5's 58.6%) - supports a genuine performance advantage that justifies the premium for long-horizon, complex tasks.

Access structure, per TrueFoundry's developer guide (June 9, 2026): API users access claude-fable-5 immediately at standard token pricing. Subscription plan users (Claude Pro, Max, Team, and seat-based Enterprise) can use Fable 5 at no extra cost through June 22, 2026. After June 23, subscription users need paid usage credits until Anthropic restores capacity. US-only inference is available at 1.1x pricing for data-residency requirements. On Claude.ai subscription plans, Fable 5 counts as 2x usage credits.

Safety classifiers introduce an important operational note: GitHub's changelog (June 9, 2026) notes that Claude Fable 5 in GitHub Copilot requires up to 30 days of prompt-and-output retention to operate the safety classifiers, with the Copilot admin policy off by default. The safety classifiers return stop_reason: "refusal" as an HTTP 200 response and are not billed if no output is generated. This is the first Anthropic model with an explicit fallback architecture - queries that trigger the classifier go to Opus 4.8 rather than being refused outright.

4. Claude Mythos 5 - The Unrestricted Twin for Glasswing Partners

Claude Mythos 5 launched simultaneously with Fable 5 on June 9, available exclusively to existing Project Glasswing partners. Both models share the same underlying weights at the same $10/$50 pricing. The distinction is access control: Mythos 5 has no safety classifiers on cybersecurity and biology queries. Per DigitalApplied's model breakdown the Mythos 5 Terminal-Bench 2.1 score of 88.0% and Health score of 66.0% both exceed Fable 5's performance on those benchmarks precisely because they are not redirected to Opus 4.8 by the classifiers.

Anthropic plans to expand Mythos 5 access in phases: first to biology researchers (timeline not specified), then to a broader cybersecurity program. The company has been explicit that the cybersecurity capabilities in the Mythos tier are powerful enough to "cause serious damage" without appropriate safeguards, per 9to5Mac's Fable 5 report (June 9, 2026). Anthropic president Daniela Amodei described Mythos as "very good at cyber warfare" in a statement this week tracked by AI Weekly.

5. Apple vs EU - Siri AI Blocked in Europe Over DMA Interoperability Standoff

The day after Apple's WWDC keynote announced a Gemini-powered Siri for the world, the EU confirmed that none of it will reach European users. The European Commission's statement on June 9, 2026 -- reported by MacRumors (June 9, 2026) -- directly contradicts Apple's characterisation of the standoff. Apple announced that Siri AI would not launch in the EU, blaming regulators for "refusing to engage constructively on proposed solutions." The Commission's account tells a different story.

Per the EU's statement (quoted in AppleInsider, June 9, 2026): "Instead of trying to find a suitable compliance solution, Apple simply made a request to the European Commission to be exempted from their interoperability obligations under the DMA -- and this for at least 18 months. That's not an option." The Digital Markets Act requires that any AI system Apple integrates into Siri must also grant equal access to competing AI assistants. Apple argues that Siri AI's agentic capabilities -- reading emails, editing files, taking cross-app actions -- are so deeply integrated that granting rivals equal access would create serious privacy and security risks.

The commercial impact: Apple's $1 billion/year Gemini licensing deal (confirmed at WWDC on June 8) now excludes the EU's estimated 450 million iPhone users. Google gets paid for global deployments only outside the EU. This directly reduces the effective value of the Apple-Google partnership by a meaningful share of the contracted revenue. It also raises the question of whether Apple will eventually develop a DMA-compliant architecture or simply accept that its AI features will be geographically fragmented. Per Archyde's DMA analysis (June 9, 2026), the EU has rejected Apple's "Trusted OS Agent" model as legally insufficient, because the DMA requires full platform access for competitors, not a sandboxed environment.

6. SpaceX IPO Eve - $135/Share Tomorrow, Demand Reaches $250 Billion

SpaceX prices its IPO tomorrow, June 11, 2026, with trading opening on Nasdaq under the ticker SPCX on June 12. As of June 9, the stock was already trading at $135.00 on pre-IPO markets, per Investing.com's live SPCX tracker (June 9, 2026). That is SpaceX's fixed IPO price per the S-1/A filed June 3, targeting 555.6 million shares raised for approximately $74.4 billion in net proceeds at a $1.75 trillion valuation -- the largest IPO in recorded history by deal size, surpassing Saudi Aramco's $29.4 billion in 2019.

The institutional demand picture, per TradingKey's market recap (June 9-10, 2026): the book is oversubscribed several times over, with total demand reportedly reaching $250 billion - more than three times the $75 billion being raised. Several large institutions have placed single orders exceeding $10 billion. Goldman Sachs is leading the 21-bank syndicate. Thirty percent of the float is allocated to retail investors through Robinhood, Fidelity, and Charles Schwab.

The key financial complication that analysts are debating, per CNBC's SpaceX IPO coverage (June 3, 2026): SpaceX is profitable only through Starlink ($11.4B revenue, $4.4B operating income in 2025). The xAI division -- maker of Grok, merged into SpaceX in February 2026 -- lost $2.4 billion in Q1 2026 alone (up from $936M YoY) while spending $7.7 billion in capex. Morningstar pegs fair value at $780 billion -- less than half the $1.75 trillion target. The American Federation of Teachers has asked the SEC to apply extraordinary scrutiny before pension money flows in. How SPCX trades on its first day will set the benchmark for Anthropic's and OpenAI's upcoming IPOs.

7. OpenAI Backs Away from Fully Autonomous AI Research - New "Tandem" Model for 2028

OpenAI is modifying its previously stated goal of a fully autonomous AI researcher by March 2028, according to new reporting tracked by LLM Stats' AI news feed (June 9-10, 2026). The company is now framing the 2028 milestone around a "tandem" between humans and AI systems rather than full autonomy. CEO Sam Altman and chief scientist Jakub Pachocki are also calling for an international body that could slow frontier AI development if the systems prove too powerful or too fast to govern responsibly.

The original timeline, announced in October 2025 and reported by TechRadar: OpenAI planned a system with "research intern" capabilities by September 2026, designed to significantly accelerate human scientific research. By March 2028, the company aimed to complete a fully autonomous AI researcher capable of running independent research projects. The updated framing replaces "fully autonomous" with "tandem" - a shift in language that carries regulatory and competitive implications.

Hot take: This pivot is strategically timed. OpenAI is in the process of filing its IPO. A company that promises to build a fully autonomous AI researcher that operates without human supervision faces a different regulatory and investor risk profile than one that promises a human-AI tandem. The 'tandem' framing also aligns with Anthropic's language around Claude's design philosophy -- supervised autonomy rather than unsupervised agency. Both companies appear to be converging on safety messaging as their IPO roadshows approach.

8. Taiwan Considers Sweeping AI Chip Export Curbs on China

Taiwan authorities are considering significantly stricter export controls on AI chip sales to all Chinese customers -- not just blacklisted entities like Huawei -- in order to align with US export control measures, per Bloomberg's exclusive report (June 9, 2026). The goal is to give Taiwan more legal tools to address the diversion of advanced hardware -- specifically AI servers equipped with Nvidia chips -- from Taiwan to China. The US has separately issued clarifications that export restrictions apply to Chinese firms outside China as well, closing a loophole that had allowed Chinese companies to buy Blackwell-generation chips through subsidiaries in third countries.

If enacted, Taiwan's curbs would dramatically tighten the supply of AI-grade semiconductors available to Chinese AI labs. China currently has Huawei's Ascend 910C chips as an alternative to Nvidia's restricted products, but the Ascend series trails Nvidia's Blackwell architecture in performance. A full Taiwan restriction on AI chip exports would narrow the performance gap between Chinese AI infrastructure and US AI infrastructure even further, slowing Chinese AI development on large training runs.

The geopolitical stakes: Taiwan is home to TSMC, the world's dominant contract chipmaker. TSMC manufactures chips for Nvidia, Apple, and effectively the entire US AI supply chain. If Taiwan aligns its export controls fully with the US framework, it creates a unified chip denial architecture that China cannot easily route around. The countermeasure risk - Beijing retaliating against Taiwan's semiconductor sector through trade restrictions or other pressure - makes this decision one of the most consequential AI-adjacent geopolitical moves of 2026.

9. Standard Bots Raises $200M - AI Robotic Arms at $1 Billion Valuation

New York-based Standard Bots has raised $200 million led by General Catalyst and Robostrategy at a $1 billion valuation, per Bloomberg's report (cited in LLM Stats, June 9-10, 2026). The company's mission: build AI-powered robotic arms in the United States, specifically targeting the reshoring of manufacturing capabilities that have historically depended on lower-cost labour markets. Standard Bots positions its platform as a way to make industrial robotics accessible to small and mid-sized manufacturers who cannot afford the capital and integration costs of traditional industrial robot deployments.

Standard Bots' robotic arms are designed to be programmed by showing the robot what to do, rather than writing code. AI handles the interpretation and generalisation from human demonstrations to a broader range of tasks. The $200 million raise at a $1 billion valuation is the largest funding round for US-based industrial robotics in 2026, and the General Catalyst lead signals that tier-one venture capital views AI-native manufacturing robotics as a near-term commercial opportunity rather than a long-horizon research bet.

The broader context: AI robotics funding in 2026 is following the pattern of AI software in 2023-2025 -- valuations are being set based on category conviction rather than revenue multiples. Standard Bots joins Figure AI (humanoid robots, backed by Microsoft and Nvidia), Physical Intelligence (Berkeley-based robot foundation model), and Wayve (autonomous driving) as beneficiaries of the institutional view that physical AI is the next trillion-dollar category. For the broader AI application landscape and use-case mapping, the AI Industry News & Trends hub tracks these funding rounds in the context of the full market.

10. Flourish Raises $500M at $2.5B for Brain-Inspired "Cortex AI"

Flourish, a New York-based neuroscience AI startup, closed a $500 million raise at a $2.5 billion valuation around June 4, 2026, according to SiliconAngle (June 4, 2026), CryptoBriefing's round analysis, and TechFundingNews (June 6, 2026). Jeff Bezos anchored the round with approximately $100 million (after nearly doubling an initial $50 million commitment), joined by Alphabet's GV, Lux Capital, and healthcare-focused Catalio Capital.

Flourish is building what it calls Cortex AI, a system designed to emulate brain function by mapping real neurons and their connections -- a field known as connectomics. The commercial thesis, per GreyJournal's Flourish analysis (June 6, 2026): a standard server-grade GPU uses about 30 times more energy than the human brain to process the same type of information. Cortex AI targets a 20-to-50-watt operating power budget -- comparable to the brain's roughly 20-watt draw during reasoning. If that efficiency target is achieved at real inference workloads, it would reshape data center economics and undercut the infrastructure rationale behind the current AI capital expenditure cycle.

The co-founders: Thomas Reardon created Internet Explorer at Microsoft, then founded CTRL-labs (brain-computer interfaces, acquired by Meta in 2019 for approximately $1 billion). Rob Williams is a former Amazon S-team executive. Greg Wayne, who heads Google's Project Astra at DeepMind, joined as a senior adviser with Demis Hassabis's blessing to spend 20% of his time at Flourish. By March 2026, Flourish had hired approximately two dozen senior neuroscientists and AI researchers.

Contrarian view: brain-inspired computing approaches have not displaced GPU-driven architecture despite years of development. The $500 million is a bet on a paradigm shift that has previously produced incremental results. But $100 million from the world's second-richest individual is not noise -- Bezos is making a deliberate hedge against the possibility that the transformer scaling approach hits a wall.

11. ChatGPT Crosses 1 Billion Monthly Active Users

ChatGPT crossed 1 billion monthly active users, making it the fastest consumer application in history to reach that milestone, per AI Weekly's alert tracking (June 2, 2026). The 900 million weekly active user figure, confirmed by OpenAI in February 2026, has since grown -- OpenAI is on pace to cross 1 billion weekly active users before the end of 2026, per TechnologyChecker's ChatGPT statistics (June 2026). The US leads with approximately 205 million users, roughly 17% of the global total. India's user base doubled in a single month after OpenAI launched ChatGPT Go at $4.50/month -- a localized pricing tier well below the $20 Plus subscription.

The 1 billion milestone has arrived nearly two years ahead of the original analyst projections from 2023. ChatGPT's growth rate has exceeded every social platform trajectory in history -- Facebook took 4.5 years to reach 1 billion. YouTube took 8 years. Instagram took 6 years. ChatGPT reached the milestone in approximately 3.5 years from launch in November 2022.

One data point that should temper the celebration: per Cloudflare Radar data tracked by TechnologyChecker.io, in May 2026 GPTBot overtook ClaudeBot to become the third-largest AI crawler on the web. GPTBot: 11.48% of all AI bot traffic. ClaudeBot: 9.73%. The order reversed from April (ClaudeBot 11.69%, GPTBot 9.84%). This suggests OpenAI is aggressively crawling the web to build out its knowledge base even as user growth drives the 1 billion headline. Publishers who are considering blocking AI crawlers in robots.txt are increasingly doing so in response to these crawl-to-referral ratios.

12. Claude Fable 5 Multi-Platform Rollout - AWS, GitHub Copilot, Foundry, Databricks, Snowflake

Claude Fable 5 launched simultaneously across multiple enterprise platforms on June 9, 2026. Available from day one: Claude API (claude-fable-5 model ID), Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Foundry -- per Anthropic's official documentation. Within hours of launch, availability expanded further: Claude Fable 5 landed on Databricks per StartupHub.ai (June 9, 2026), and Snowflake Cortex AI added Fable 5 to its model library on the same day. GitHub Copilot integration was confirmed in GitHub's changelog with the 30-day data retention requirement for safety classifiers.

The simultaneous multi-platform launch reflects Anthropic's enterprise distribution strategy: Claude is the only frontier AI model available across all three major cloud platforms (AWS Bedrock, Google Cloud Vertex AI, and Microsoft Azure Foundry). By ensuring Fable 5 lands on all three on day one, Anthropic ensures enterprise IT teams do not face a procurement delay waiting for their preferred cloud provider to certify the new model. For implementation guidance on integrating Fable 5 into existing Claude API workflows -- including handling the new stop_reason: "refusal" response type and optimising for prompt caching -- the gen-ai-experiments cookbook repository covers production deployment patterns for Anthropic models across cloud environments.

13. Goldman Sachs CEO: Markets in "Greed Mode" for AI IPOs

Goldman Sachs CEO David Solomon described markets as being in "Greed Mode" as SpaceX, Anthropic, and OpenAI prepare their historic AI IPO wave, per reporting tracked by AI Weekly (June 2026 edition). Solomon's characterisation reflects the oversubscription data: SpaceX's IPO has generated $250 billion in demand against a $75 billion raise. Goldman Sachs projects 2026 IPO proceeds could reach $160 billion -- a quadrupling from 2025 -- driven almost entirely by SpaceX, Anthropic, and OpenAI.

The context: Goldman Sachs is leading SpaceX's 21-bank IPO syndicate, giving Solomon a closer view of demand than almost any other external observer. When the CEO of the lead underwriter describes market sentiment as 'greed mode,' it carries more weight than an analyst's view. The phrase also carries a cautionary note -- greed mode describes the late stage of a bull market cycle, not the beginning.

The 'greed mode' designation has historical precedent in technology IPOs. The dot-com era of 1999-2000 was characterised by institutional investors abandoning fundamental valuation frameworks for narrative-driven pricing. The 2021 SPAC boom showed similar dynamics. In both cases, 'greed mode' correctly described the sentiment before corrections that wiped out significant portions of those valuations. Whether the 2026 AI IPO wave represents a fundamentally different situation -- because these companies have real revenue, unlike many dot-com era IPOs -- is the central debate in institutional AI investing right now.

14. Microsoft Lays Off 200-400 Azure Employees in China - Third Round of Cuts

Microsoft has laid off between 200 and 400 Azure unit employees in Beijing and Shanghai, per South China Morning Post reporting (cited in LLM Stats, June 9-10, 2026). This is at least Microsoft's third round of downsizing in China in two years. The Azure China workforce reduction follows Microsoft's broader strategic repositioning away from mainland China cloud operations amid the intensifying US-China technology competition and export control environment.

The timing connects directly to the Taiwan AI chip export curbs story (Story 8): if Taiwan enacts stricter restrictions on AI chip sales to China aligned with US measures, Microsoft's operational footprint in China becomes increasingly difficult to sustain at scale. Azure relies on the same Nvidia GPU ecosystem that is being restricted. A Microsoft Azure China operation that cannot access frontier AI hardware cannot serve enterprise customers who need frontier AI capabilities.

The broader pattern: Google's DeepMind and other major US AI labs have reduced China-based headcount significantly in 2025-2026. The AI industry's centre of gravity is concentrating in the US, UK, and select allied nations. China is building its own AI ecosystem with domestic hardware (Huawei Ascend) and domestic models (Qwen, DeepSeek) precisely because the US-aligned AI stack is becoming unavailable to it at the frontier level.

15. AI Web Crawler Wars - ClaudeBot at 11,122:1 Ratio vs Google's 5:1

The asymmetry between how much AI crawlers take from the web and how much they send back is emerging as a major publisher policy issue. Per Cloudflare Radar data analysed by TechnologyChecker.io (June 2026 edition): Anthropic's ClaudeBot crawled 11,122 pages for every 1 human visit Anthropic sent back to the web in the week of May 25 to June 1, 2026. OpenAI's GPTBot ratio: 857:1, down from 1,252:1. Perplexity moved the wrong way, from 95:1 to 190:1. Google's traditional Googlebot: 5:1. ClaudeBot improved from 13,528:1 in April, but remains the worst ratio of any major operator by a wide margin.

This asymmetry is the structural reason publishers have started blocking AI crawlers in robots.txt. From a publisher's perspective, ClaudeBot arriving 11,122 times for every 1 human referral it returns represents a net drain of content without commercial compensation. Google's traditional search model -- which functions at 5:1 -- compensates publishers with discovery traffic. AI crawlers that consume content to train models or build retrieval systems without proportionate traffic return create a different economic relationship.

Hot take: The 11,122:1 ratio for ClaudeBot reflects how Anthropic's business model differs from Google's. Google's PageRank algorithm is built on sending traffic back to publishers -- that return is the commercial logic that justifies crawling. Anthropic's Claude doesn't have a search discovery product that sends equivalent traffic. As AI companies move toward more web-grounded responses (Fable 5's knowledge cutoff and retrieval capabilities), the expected payback is increasing referral traffic. If that doesn't materialise, the robots.txt blocking movement will accelerate, degrading the training data quality available to all AI companies.

16. The Fable 5 Access Window - What Every Paid Claude Subscriber Needs to Know Before June 23

There are 13 days left on the free Fable 5 access window for paid Claude subscribers. Per Anthropic's official announcement and confirmed by PCWorld (June 9, 2026) and LushBinary's developer guide: Claude Pro, Max, Team, and seat-based Enterprise subscribers can use Fable 5 at no extra cost through June 22, 2026. After June 23, Fable 5 access requires paid usage credits. Anthropic's stated intent is to restore Fable 5 as a standard part of subscription plans when capacity allows -- no committed date has been given.

Practical implications for different user types. API developers: the June 23 date does not apply to you -- you pay per token regardless and Fable 5 is available immediately. This is the better access model for teams doing systematic evaluation or production deployment. Claude Pro subscribers ($20/month): you have until June 22 to run your most demanding tasks against Fable 5 -- large codebase reviews, multi-step research, complex document analysis. The $10/$50 per-million-token pricing starts on June 23 if you continue using it.

Claude Team and Enterprise admins: note the 2x usage credit consumption rate and the 30-day data retention requirement for safety classifiers in GitHub Copilot. The latter may require IT security review before enabling Fable 5 in your Copilot environment. Given that Fable 5 scores 80.3% on SWE-Bench Pro versus Opus 4.8's 69.2% -- an 11-point gap on the benchmark that best predicts real-world software engineering quality -- the June 9-22 window is worth treating as a structured evaluation sprint for engineering teams considering whether to switch their primary model. For a practical guide to structuring that evaluation -- what tasks to test, how to measure output quality versus cost, and when Opus 4.8 remains the better choice -- the gen-ai-experiments repository has model comparison notebooks and evaluation frameworks you can run against your own tasks.

Frequently Asked Questions

What is Claude Fable 5 and how does it differ from Claude Opus 4.8?

Claude Fable 5 is Anthropic's first publicly available Mythos-class AI model, launched June 9, 2026. It is the same underlying model as Claude Mythos 5 but includes safety classifiers that redirect cybersecurity and biology queries to Claude Opus 4.8. It sits above the Opus class in capability: SWE-Bench Pro 80.3% vs Opus 4.8's 69.2%, GDPval-AA 1932 vs 1890, and Terminal-Bench 2.1 88.0% (Mythos 5). Priced at $10/$50 per million input/output tokens -- double Opus 4.8 ($5/$25). Designed for long-horizon, complex tasks: large code migrations, multi-day agentic sessions, deep research. Available via Claude API, Amazon Bedrock, Google Cloud Vertex AI, Microsoft Foundry, GitHub Copilot, Databricks, and Snowflake. Source: Anthropic.com/news/claude-fable-5-mythos-5.

Why is Apple blocking Siri AI in Europe?

Apple announced that iOS 27's rebuilt Siri AI (powered by Google Gemini) will not be available to EU users. Apple blames the Digital Markets Act's interoperability requirements, saying they would demand rivals be given "nearly unlimited access" to a user's device. The EU Commission disputes this characterisation, stating that Apple never proposed a compliant solution -- it simply requested an 18-month blanket exemption from DMA obligations, which the Commission rejected as "not an option." The DMA requires that if Apple integrates Siri AI into iOS, it must grant equal platform access to competing AI assistants. Source: MacRumors, June 9, 2026; AppleInsider, June 9, 2026.

When does SpaceX start trading on Nasdaq?

SpaceX prices its IPO on June 11, 2026, with trading opening on Nasdaq under the ticker SPCX on June 12, 2026. The fixed IPO price is $135 per share. SpaceX is selling approximately 555.6 million Class A shares to raise about $74.4 billion in net proceeds at a $1.75 trillion valuation -- the largest IPO in recorded history by deal size. Institutional demand has reportedly reached $250 billion, multiple times the deal size. Goldman Sachs is leading the 21-bank underwriting syndicate. Source: Investing.com SPCX tracker (June 9, 2026); Polymarket; CNBC.

What is Flourish AI and what is Cortex AI?

Flourish is a New York-based AI startup that raised $500 million at a $2.5 billion valuation in early June 2026, backed by Jeff Bezos (~$100M), Alphabet's GV, Lux Capital, and Catalio Capital. The company is building Cortex AI -- a brain-inspired AI system based on connectomics (mapping real neurons and connections). The goal is to achieve efficient inference at 20-50 watts of power (comparable to the human brain), versus a standard server-grade GPU's approximately 600 watts. Co-founders: Thomas Reardon (created Internet Explorer, sold CTRL-labs to Meta for ~$1B) and Rob Williams (former Amazon S-team). Source: SiliconAngle; CryptoBriefing.

Did ChatGPT really reach 1 billion users?

ChatGPT crossed 1 billion monthly active users in June 2026, making it the fastest consumer app in history to reach that milestone. OpenAI confirmed 900 million weekly active users in February 2026. At the current trajectory, ChatGPT is on pace to cross 1 billion weekly active users before the end of 2026. The US leads with approximately 205 million users; India's base doubled in one month after ChatGPT Go launched at $4.50/month. ChatGPT also processes over 2.5 billion queries per day. Source: DemandSage ChatGPT statistics (June 2026); AI Weekly alert (June 2, 2026).

Why is Taiwan considering restricting AI chip exports to China?

Taiwan is considering sweeping export controls on AI chip sales to all Chinese customers (not just blacklisted entities like Huawei) to align with US export control measures. The primary concern is semiconductor smuggling -- AI servers equipped with Nvidia chips are being diverted from Taiwan to China through third-party channels. Bloomberg (June 9, 2026) reported that Taiwan authorities want more legal tools to address this diversion. The US has separately clarified that export restrictions apply to Chinese-headquartered firms outside China as well. Source: Bloomberg (June 9, 2026).

Is Claude Fable 5 free on Claude Pro subscriptions?

Claude Fable 5 is included at no extra cost for Claude Pro, Max, Team, and seat-based Enterprise subscribers from June 9 through June 22, 2026. After June 23, paid usage credits are required for Fable 5 access until Anthropic restores sufficient capacity to include it as a standard plan feature. Fable 5 counts as 2x usage credits on subscription plans. API users pay the standard token pricing ($10/$50 per million tokens) with no time-limited window. Source: Anthropic official announcement; PCWorld (June 9, 2026); TrueFoundry developer guide.

Reference Links

Claude Fable 5 and Claude Mythos 5

• Anthropic -- Claude Fable 5 and Claude Mythos 5 (Official Announcement, June 9, 2026)�/u��u�PCWorld -- Claude Fable 5: Too Dangerous AI Model Is Finally Public (June 9, 2026)�/u��u�9to5Mac -- Anthropic Releases Public Mythos-Class AI Model Called Claude Fable (June 9, 2026)�/u��u�Cryptopolitan -- Anthropic Launches Claude Fable 5, Its Most Capable Publicly Available AI Model (June 9, 2026)�/u��u�Vellum.ai -- Claude Fable 5 and Mythos 5 Benchmarks Explained (June 9, 2026)�/u��u�TrueFoundry -- Claude Fable 5: API, Benchmarks, Pricing & How to Use It (June 9, 2026)�/u��u�DigitalApplied -- Claude Fable 5 and Mythos 5: The Frontier, Split in Two (June 9, 2026)�/u��u�Finout.io -- Claude Fable 5 and Mythos 5 Pricing and Benchmark Comparison (June 9, 2026)�/u��u�LushBinary -- Claude Fable 5 Developer Guide (June 9, 2026)�/u��u�Let's Data Science -- Anthropic Releases Claude Fable 5 for Broad Use (June 9, 2026)�/u��u�StartupHub.ai -- Claude Fable 5 Arrives on Databricks (June 9, 2026)�/u��u�GuruFocus -- Microsoft Launches Claude Fable 5 for Enhanced AI Capabilities (June 9, 2026)�/u��u�MacRumors -- EU Says Decision Not to Launch Siri AI in Europe Is Apple's Alone (June 9, 2026)�/u��u�AppleInsider -- Siri AI Is Trapped in Apple's Escalating Cold War with the EU (June 9, 2026)�/u��u�Archyde -- EU Rejects Apple's Siri AI Exemption: DMA Rules Non-Negotiable (June 9, 2026)�/u��u�AppMust -- Apple's Siri AI Stand-Off with Europe Just Escalated (June 9, 2026)�/u��u�Investing.com -- SpaceX (SPCX) Pre-IPO Trading at $135.00 (June 9, 2026)�/u��u�TradingKey -- SpaceX IPO Subscriptions Surge to $250 Billion (June 9-10, 2026)�/u��u�TradingKey -- SpaceX IPO Receives Multi-Fold Institutional Over-Subscriptions (June 9, 2026)�/u��u�Polymarket -- SpaceX IPO by [date] Prediction (June 2026)�/u��u�CNBC -- SpaceX Targets Fixed $135 IPO Price for Roadshow (June 3, 2026)�/u��u�DecodetheFuture -- AI IPOs 2026: Anthropic vs OpenAI vs SpaceX Analysis�/u��u�LLM Stats -- LLM News Today (June 2026) -- OpenAI Tandem Reference (June 9-10, 2026)�/u��u�The Decoder -- OpenAI Targets Full-Scale Autonomous AI Researcher by Early 2028 (Original October 2025)�/u��u�Bloomberg -- Taiwan Weighs Tighter AI Chip Export Controls Targeting China to Align with US (June 9, 2026)�/u��u�StartupHub.ai -- Taiwan Eyes China AI Chip Sales Curbs; OpenAI IPO Looms (June 9, 2026)�/u��u�LLM Stats -- Bloomberg Report: Standard Bots Raises $200M at $1B Valuation (June 9-10, 2026)�/u��u�CryptoBriefing -- Flourish Secures $500M from Jeff Bezos for Brain-Inspired AI (June 2026)�/u��u�SiliconAngle -- AI Startup Flourish Reportedly Raises $500M Round Backed by Jeff Bezos (June 4, 2026)�/u��u�TechFundingNews -- Bezos Commits Close to $100M to Flourish (June 6, 2026)�/u��u�GreyJournal -- Bezos Backs Flourish, a $500M Brain-Inspired AI Startup (June 2026)�/u��u�TechTimes -- Jeff Bezos Bets on Flourish $500M Startup (June 6, 2026)�/u��u�AI Weekly -- Flourish Closes $500M Round to Build Brain-Inspired AI (June 2026)�/u��u�DemandSage -- ChatGPT Statistics June 2026: 1 Billion Monthly Active Users�/u��u�TechnologyChecker.io -- ChatGPT Statistics 2026: 900M Users, Cloudflare Crawl Data (June 2026)�/u��u�AI Weekly -- Alert: ChatGPT Crosses 1 Billion Monthly Active Users (June 2, 2026)�/u��u�AI Weekly -- Goldman Sachs CEO David Solomon: Markets in "Greed Mode" (June 2026)�/u��u�LLM Stats -- South China Morning Post: Microsoft Lays Off 200-400 Azure Employees in China (June 9-10, 2026)�/u��u�TechnologyChecker.io -- ChatGPT Statistics 2026: ClaudeBot and GPTBot Cloudflare Crawl Ratios (June 2026)

Recommended Blogs

• AI News Today -- June 8, 2026: 16 Biggest Stories (WWDC, Claude on iPhone, Microsoft Foundry, EU AI Act)�/u��u�AI News Today -- June 7, 2026: 16 Biggest Stories (Trump+Sanders AI Ownership, Grok Build, Anthropic Brake Pedal)�/u��u�AI News Today -- June 6, 2026: 16 Biggest Stories (ChatGPT Dreaming V3, Anthropic IPO, MAI Models)�/u��u�AI News Today -- June 5, 2026: Great American AI Act, NVIDIA RTX Spark, Glasswing Expansion�/u��u�AI Industry News & Trends Hub -- Complete BFWAI Daily AI News Coverage�/u��u�Claude AI Complete Hub -- All Anthropic Models: Fable 5, Opus 4.8, Sonnet 4.6, Mythos, Code, Glasswing�/u��u�AI Coding Tools Hub -- Claude Fable 5 vs GPT-5.5 vs Grok Build vs GitHub Copilot (2026)

Follow Build Fast with AI for daily AI news, model reviews, and builder-first implementation guides. New post every morning at 8 a.m. IST.