Regulators just did to Google what no competitor could. The European Commission ordered Google to open Android to rival AI assistants and share its search data with competing AI developers, binding decisions that reshape who gets to reach two billion phones. It caps a brutal stretch for Google, whose Gemini 3.5 Pro reportedly missed its target a third time. Meanwhile Oracle is cutting up to 30,000 jobs to fund the $500 billion Stargate buildout, and SAP put over a billion euros behind a European frontier lab betting on something other than chatbots.
Here are the 16 stories that matter for July 20, 2026, with the numbers, dates, and honest caveats. For running coverage of every release this month, bookmark our AI industry news and trends hub.
1. The EU Orders Google to Open Android and Share Search Data With AI Rivals
The European Commission adopted binding requirements under the Digital Markets Act ordering Google to open Android to rival AI assistants and to share portions of its search data with competitors, including AI developers. Under the decision, eligible third-party assistants gain voice activation and cross-app capabilities across 11 Android feature groups, subject to certification and user consent, while Google must make anonymized ranking, query, click, and view data available on fair, reasonable, and nondiscriminatory terms. Search data sharing begins in January 2027, with Android interoperability due by July 2027.
This is the most consequential regulatory action in AI this year, because it attacks the two assets that make Google nearly unbeatable: default placement on billions of Android devices, and two decades of search behavior data that no competitor can replicate. Letting a rival assistant activate by voice and work across apps on Android removes the structural advantage Gemini enjoys by simply being preinstalled. Handing over anonymized search data gives AI developers training and ranking signal they could never otherwise buy. Google's president of global affairs Kent Walker pushed back, arguing the decisions risk undermining privacy and security guardrails for millions of Europeans.
The timing lands with unusual force. Google spent this month failing to ship its flagship model, and now regulators are prying open the distribution moat that was supposed to compensate. For every AI company that is not Google, this is the best news of the month: a legal path onto Android and access to search signal, arriving right as Google looks vulnerable. My take: distribution has been the quiet answer to why Google survives bad model weeks, and Brussels just put a timer on that answer. Watch which assistants apply for certification first.
2. Gemini 3.5 Pro Misses a Third Time as Google Weighs a Stopgap Release
Gemini 3.5 Pro reportedly missed its July 17 target, marking the third slip for Google's flagship, and the company is now said to be exploring a stopgap Gemini 3.6 Flash release to put something in market while the Pro model gets fixed. Google has still published no official model card, pricing, or benchmarks, so every circulating claim remains unconfirmed. Alphabet shares fell about 4 percent on the delay reports, as we covered in our July 18 AI news recap.
A third miss changes the nature of the problem. One delay is engineering discipline; three suggests something structural, whether in the training run, the evaluation bar Google has set for itself, or both. The reported stopgap is the detail worth watching, because shipping a Flash-tier model to fill a Pro-tier gap is a tacit admission that the flagship is not close. It would give Google a fresh release to point at while buying months, but it would also confirm to enterprise buyers that the top-end Gemini they have been waiting on is not imminent.
The competitive cost compounds daily. Enterprises evaluating frontier models this quarter are choosing among GPT-5.6, Claude, Grok 4.5, and now Kimi K3, and every week Gemini is absent is a week those contracts get signed elsewhere. Where every shipped model actually ranks is tracked on our best AI models July 2026 leaderboard. My honest read: Google's research depth is real and this is recoverable, but the company needs to either ship something credible or say publicly what is happening, because silence plus slippage is the worst combination for enterprise trust.
3. Kimi K3 Rattles the US AI Industry as the Open-Weight Shock Lands
Moonshot AI's Kimi K3 stunned the US technology industry over the weekend, setting off fresh debate about the China and US AI rivalry, after the 2.8-trillion-parameter open model took the top spot on a major coding leaderboard days earlier. The reaction story is now as significant as the launch itself, with American labs and investors publicly reassessing how far ahead the closed frontier really is.
What makes K3 land differently from previous Chinese releases is the combination of scale, benchmark position, and the promise of free weights on July 27. Earlier Chinese models competed on price; K3 competed on capability and won on a coding leaderboard against Claude Fable 5, then announced it would give the weights away. That sequence removes the two comfortable arguments US labs have used, that open models trail on quality and that Chinese models are cheap substitutes rather than genuine frontier systems. For developers, our AI coding tools hub is tracking where K3 Max actually holds up in production work versus where the benchmark flatters it.
The honest caveat is that leaderboard wins are narrow. K3 ranks around ninth on general chat, so it is a coding and agent specialist rather than an all-around frontier replacement, and independent evaluation across varied workloads is still thin. But specialists are exactly what enterprises deploy for high-volume coding, and a free specialist that beats paid generalists on the task you care about is a genuinely difficult thing to argue against on a budget review. The July 27 weight release is the moment this stops being a benchmark story and becomes a procurement story.
4. Oracle Cuts Up to 30,000 Jobs to Fund the $500 Billion Stargate Buildout
Oracle is cutting up to 30,000 employees, roughly 18 percent of its global workforce, to free an estimated $8 to $10 billion in annual cash flow for AI infrastructure, in the largest workforce reduction in the company's history. The cuts fund Oracle's role in Stargate, the $500 billion AI infrastructure initiative with OpenAI and SoftBank, anchored by a $300 billion five-year cloud contract with OpenAI that is expected to generate roughly $30 billion a year from 4.5 gigawatts of Oracle-built data center capacity.
The internal allocation tells the story better than the headline number. The reductions hit Oracle Health, cloud infrastructure, and consulting hardest while sparing the teams building Stargate data centers, which Oracle is racing to staff. That is a company converting itself, one department at a time, into an AI infrastructure provider, and financing the transformation with the salaries of the businesses it is deprioritizing. It is the clearest example yet of how the AI capital expenditure boom is actually being funded: not entirely with new money, but by redirecting cash from existing operations.
The strategic bet is enormous and concentrated. Oracle has tied its future to a single customer relationship, and a $300 billion contract with OpenAI means Oracle's returns depend on OpenAI's growth, its ability to pay, and the durability of a company currently facing an Apple lawsuit and publisher litigation ahead of an IPO. My take: this is the most leveraged bet any large enterprise vendor has made on AI, and the human cost of 30,000 jobs makes it the starkest illustration of what the buildout actually requires. If Stargate delivers, Oracle reinvents itself. If it does not, the cuts bought nothing.
5. Microsoft's Project Perception Takes On Anthropic in AI Security
Microsoft is preparing Project Perception, an AI cybersecurity platform that finds and fixes software vulnerabilities using models from Microsoft, OpenAI, and Anthropic together, positioned as a lower-cost alternative to Anthropic's Mythos-class security offering. The system looks across a company's code, cloud infrastructure, and endpoints, identifies exploitable weaknesses, explains their impact, and proposes concrete fixes. Microsoft has not publicly confirmed availability, pricing, or customer eligibility.
The architectural detail is the genuinely interesting part, and it is a pattern more teams should copy. Project Perception uses an orchestration layer that routes each task to the best-fit model rather than sending everything to the most powerful and most expensive one. A cheap model handles inventory checks, log parsing, and initial triage of common vulnerability types, while a frontier model gets called only when the system needs to reason through a complex exploit chain or write a remediation plan touching production. That routing is what makes continuous, always-on vulnerability scanning affordable instead of a budget line nobody approves.
This is the good-news story of the week for defenders. The cost of running frontier models against an entire codebase has been the main reason continuous AI security auditing stayed theoretical, and smart routing plausibly solves it. If you are building agent systems with mixed-cost model routing, the orchestration patterns in our open-source Gen AI cookbooks cover the same technique. My take: Microsoft using Anthropic's models inside a product designed to compete with Anthropic is peak 2026, and the cost engineering matters more than the rivalry.
Don't just use ChatGPT. Learn to build custom LLM agents, RAG pipelines, and full-stack Agentic AI apps in our intensive 6-week program.
6. SAP Completes the Prior Labs Deal With Over 1 Billion Euros
SAP completed its acquisition of Prior Labs, the Freiburg-based pioneer of tabular foundation models, and committed to investing more than 1 billion euros over four years to scale it into a globally leading frontier AI lab. Prior Labs will continue operating as an independent entity. The startup, founded by Frank Hutter, Noah Hollmann, and Sauraj Gambhir roughly 18 months ago, built the TabPFN model series that was published in Nature and set the state of the art on tabular benchmarks across hundreds of independent academic studies.
The reasoning behind the deal is refreshingly contrarian. SAP concluded that the biggest untapped opportunity in enterprise AI was not large language models but AI purpose-built for the structured data that actually runs businesses: the tables, ledgers, inventories, and transaction records sitting in enterprise databases. Language models handle documents and chat well and handle spreadsheets poorly, and SAP sits on more enterprise structured data than almost anyone. Buying the leading tabular model lab and funding it at a billion euros is a bet that this category becomes as important as chatbots for actual business value.
It is also a meaningful European AI story at a moment when the continent is usually cast as regulating rather than building. An 18-month-old German startup with a Nature paper being scaled into a frontier lab with a billion euros behind it is the kind of outcome European technology policy has been trying to manufacture for a decade. My take: this is the most interesting acquisition of the month precisely because it ignores the chatbot race entirely, and I suspect tabular models will quietly deliver more measurable enterprise ROI than another point of benchmark gain.
7. Tabular Foundation Models: The Frontier That Is Not a Chatbot
Tabular foundation models are AI systems pretrained on structured, table-shaped data rather than text, designed to make predictions on spreadsheets, databases, and business records the way language models make predictions on words. Prior Labs' TabPFN series demonstrated that a single pretrained model can outperform traditional machine learning approaches on tabular benchmarks without task-specific training, a result strong enough to publish in Nature and now strong enough to justify SAP's billion-euro commitment.
The practical importance is easy to underestimate because it is unglamorous. Most of the data that businesses actually run on is tabular: sales records, supply chain tables, financial ledgers, sensor logs, customer databases. Companies have spent years trying to force this data through language models with mixed results, because a model trained on prose is not naturally suited to a million-row table. A foundation model built specifically for that shape of data can do forecasting, anomaly detection, and prediction directly, without the elaborate feature engineering that traditional approaches require.
For builders, this is a category worth understanding now rather than in two years. If your problem is predicting a number from a table, a tabular foundation model may beat both a fine-tuned language model and a hand-built gradient boosting pipeline, with far less setup. My take: the AI conversation has been so dominated by chatbots that an entire adjacent frontier got very little attention, and SAP just paid a billion euros to say that was a mistake. Expect competitors to notice quickly.
8. Claude Fable 5's Free Access Expires and Forces a Decision
Anthropic's free access window for Claude Fable 5 expires at 11:59 PM Pacific on Sunday July 19, forcing a decision point on what comes next: an Opus 5 release, a fourth extension of free access, or a shift to a credits-based model. Fable 5 has been one of the strongest models available during the free window, and its expiry lands the same weekend Kimi K3 arrived promising free weights on July 27.
The timing puts Anthropic in an awkward position it did not choose. Ending free access to a flagship model in the same week a rival open model tops the coding charts and announces free weights is a difficult contrast to manage, even if the underlying economics are entirely reasonable. Free access windows exist to drive adoption and gather feedback, and they always end. But the competitive backdrop has changed since this one started, and users moving off free Fable 5 now have a genuinely capable free alternative arriving within days.
The likeliest read is that Anthropic uses this moment for an Opus 5 announcement, which would reset the conversation on its terms rather than on Moonshot's. The company had the best all-around week of any lab, with a confidential IPO filing, the top safety grade, and the reported Karpathy hire, so it is negotiating from strength. My take: whatever Anthropic announces next will be read as its answer to Kimi K3, fairly or not, and that is exactly the sort of pressure the open-weight offensive was designed to create.
9. AI Text Detectors Fail When Models Imitate a Writer's Style
Epoch AI tested three leading AI text detectors, Pangram, GPTZero, and Originality.ai, against text generated in imitation of a specific author's style, and found that up to 18 percent of AI-generated passages went undetected. Scientific writing proved particularly vulnerable, with detectors struggling most on the formal, structured prose common in academic work.
An 18 percent miss rate matters enormously in the places these tools are actually deployed. Universities use detectors to police academic integrity, publishers use them to screen submissions, and hiring managers use them to evaluate written work, often treating a detector result as decisive. A tool that misses nearly one in five AI passages when someone applies a simple style-imitation prompt is not a reliable basis for consequential decisions, and the vulnerability in scientific writing is especially concerning given how much academic screening now relies on automated detection.
The deeper problem is that detection is losing an asymmetric race. Making a model imitate a writing style is trivially easy, requiring only a prompt, while detecting the result is a hard statistical problem that gets harder as models improve. My take: institutions should stop treating AI detectors as evidence and start treating them as weak signals at best. The honest path forward is redesigning assessment around process and verification rather than trying to catch outputs, because the catching approach is not winning and the study numbers show why.
The only comprehensive program designed to take you from basic prompting to building interactive Artifacts, custom integrations, and deploying production-ready code with Claude Code.
10. AI Radiology Models Are Confidently Wrong, a New Benchmark Finds
A new benchmark called RadLE 2.0 tested AI models on radiology tasks and found they frequently deliver wrong findings with full confidence, reading X-rays and producing incorrect diagnoses without any signal of uncertainty. The confidently-wrong failure mode is the specific risk, because a hedged wrong answer invites a second opinion while a confident wrong answer does not.
This lands with real weight given how fast AI is moving into healthcare. This month alone brought Neko Health's $700 million round for AI-analyzed body scans, Hemispheric's $52 million for brain-activity AI, and the US government deploying ChatGPT to audit Medicare and Medicaid data. Each of those depends on AI outputs being trustworthy or at least appropriately uncertain, and a benchmark showing models are confidently wrong in radiology is a caution that applies across all of them. Miscalibrated confidence is arguably more dangerous than raw error rate, because it defeats the human review that is supposed to catch mistakes.
The constructive read is that benchmarks like RadLE 2.0 are exactly what the field needs, since you cannot fix what you do not measure, and publishing failure modes is how medical AI earns the trust it will eventually deserve. My take: any AI system deployed in a clinical setting should be required to express calibrated uncertainty, and if it cannot, a human should never see its output as a conclusion. The technology is genuinely promising here; the deployment discipline has not caught up.
11. The World AI Conference Closes as WAICO Takes Shape
The 2026 World AI Conference in Shanghai closes today, July 20, after four days that included Xi Jinping's first-ever keynote and the launch of the World Artificial Intelligence Cooperation Organization with 29 founding countries including Pakistan, Russia, and Kazakhstan. The conference ran more than 140 forums with over 1,100 exhibitors, and Huawei used the show floor to demonstrate its Atlas 950 SuperPoD domestic computing system.
What matters now is what survives the closing ceremony. A governance organization announced with fanfare either becomes an institution with staff, standards, and a work program, or it becomes a communique nobody references again. The signals to watch are whether WAICO publishes a founding charter, names leadership, sets a meeting calendar, and attracts members beyond the initial 29, particularly any European or Global South economies not already aligned with Beijing. Xi paired the launch with a strong endorsement of open-source AI and pledges of assistance to developing countries, which is the recruitment pitch.
The Western response remains the open question, and it is conspicuously absent. Google DeepMind CEO Demis Hassabis called for an international watchdog and a US-led coalition the same week, which reads as an admission that no such body exists while China's now does. My take: institutions are built slowly and matter for decades, and the side that shows up with a charter and a headquarters usually shapes the rules. Whatever one thinks of WAICO's motives, arriving first with structure is a real advantage.
12. Google's Brutal Week: A Delay, an EU Order, and a Stock Drop
Put the week together and Google absorbed three distinct blows in seven days: Gemini 3.5 Pro missed its target a third time, the European Commission ordered it to open Android to rival AI assistants and share search data, and Alphabet shares fell about 4 percent. Each is survivable alone. Arriving together, they hit both halves of Google's AI strategy at once, the model and the distribution that was supposed to compensate for the model.
The fair counterweight is that Google shipped real things this week that got buried. It renamed NotebookLM to Gemini Notebook and gave it a secure cloud computer that runs code inside notebooks, serving more than 30 million users and 600,000 organizations, and it expanded AI Mode in Search with Instacart, Canva, and YouTube Music integrations that turn search into completed actions. Those are meaningful product wins from a company with distribution nobody else can match. They simply could not compete for attention against a flagship delay and a regulatory order.
My honest take: Google is not in decline, it is in a bad stretch with an unforgiving news cycle, and the EU order is the more serious long-term problem because a model can be fixed while a structural remedy lasts. The company still has the deepest research bench in AI, the most-used products on the internet, and its own silicon. But it needs to ship a credible frontier model soon, because the narrative is hardening and enterprise contracts signed elsewhere this quarter do not come back next quarter.
13. AI Security Becomes a Two-Horse Race
With Microsoft preparing Project Perception to compete against Anthropic's Mythos-class security offering, AI-powered vulnerability detection is consolidating into a genuine two-horse race between the company with the largest enterprise distribution and the company with the strongest security-specific model. Anthropic's Project Glasswing already expanded to 150 organizations across 15 countries, and Microsoft is countering with multi-model routing and its own vast install base.
The competition is arriving because the problem is real and urgent. Microsoft's own July Patch Tuesday fixed a record 570 vulnerabilities with AI assistance, AI security acquisitions tripled from 10 last year to 29 in the first half of 2026, and CISA warned that autonomous agents are opening new gaps in identity and access management. Every agent granted credentials is a new attack surface, and defenders need tooling that operates at the same speed as AI-assisted attackers. Two well-resourced competitors racing on cost and coverage is genuinely good for the organizations that need this.
The differentiator will likely be economics rather than raw capability. Anthropic leads on model quality for security reasoning, while Microsoft's routing architecture attacks the cost problem that keeps continuous scanning off most budgets, plus it can bundle into existing enterprise agreements. My take: this is one of the healthiest competitive dynamics in AI right now, because both companies are pushing toward making machine-speed defense affordable, and the alternative to that competition is a world where only the largest enterprises can afford to defend themselves.
14. The Open-Weight Countdown: DeepSeek July 24, Kimi K3 July 27
Two dates now anchor the rest of July. DeepSeek's V4 stable release lands July 24, ending the preview-build churn that has kept cautious enterprises from moving production workloads onto it, and Kimi K3's open weights go free on July 27, putting a model that just topped a coding leaderboard into anyone's hands. Together they represent the most concentrated open-weight release window the industry has seen.
The commercial stakes are straightforward. DeepSeek's roughly $0.44 per million output tokens is already the price floor the industry gets measured against, and a stable release removes the last technical objection to adopting it. K3's weights arriving three days later means enterprises can run a top-tier coding model on their own infrastructure with no per-token cost at all. For any organization spending heavily on frontier API calls for high-volume coding or agent workloads, the last week of July is the moment to run a serious evaluation.
The practical advice for teams is to treat these dates as a forcing function rather than a headline. Run your actual workloads against stable V4, K3 Max, and your current closed model, measure quality and total cost including the infrastructure to self-host, and let the numbers decide rather than the leaderboards. My take: the honest answer will be task-dependent, with closed models still winning the hardest reasoning while open models win high-volume routine work, and the teams that build routing between them will spend far less than teams that pick one.
15. What Oracle's Cuts Reveal About AI's Real Economics
Oracle eliminating up to 30,000 jobs to free $8 to $10 billion a year for data centers is the most honest accounting anyone has published of what the AI buildout costs. The capital for gigawatt-scale infrastructure is not appearing from nowhere; it is being extracted from existing business lines, headcount, and the operations that funded the company before AI became the priority. Oracle simply did it openly enough to count.
The pattern repeats across the industry once you look for it. Meta is spending $125 to $145 billion this year and committed $50 billion to a single Louisiana site while reorganizing around compute. TSMC raised capital spending guidance twice. Microsoft, Amazon, and Google are all redirecting enormous cash flows toward silicon and power. The difference is that most of these companies fund it from growing revenue, while Oracle is funding it from a restructuring, which makes the trade-off visible in a way the others avoid.
The risk in Oracle's version is concentration. Its bet rests on a $300 billion, five-year contract with a single customer, OpenAI, whose own position includes an Apple lawsuit, publisher litigation, and an unproven path to profitability at the scale the contract assumes. My take: the AI infrastructure boom is real and the demand signals from TSMC confirm it, but Oracle has taken the least diversified route available, and 30,000 people paid for that choice. It deserves to be remembered as a data point about cost, not just a strategy headline.
16. What to Watch This Week
Four things are already scheduled and follow directly from this weekend. The World AI Conference closes today in Shanghai, and whether WAICO publishes a charter, names leadership, or adds members will show if it is an institution or an announcement. DeepSeek V4's stable release lands July 24. Kimi K3's open weights go free July 27. And Google faces the compounding question of whether to ship a stopgap Gemini 3.6 Flash or hold out for a Pro model that clears its own bar.
Two slower storylines are worth tracking underneath the dated events. Anthropic's IPO process moves forward after the confidential S-1, and any leak about timing or valuation will move the entire sector's comparables. And the EU's Google remedies begin a long implementation runway toward January and July 2027, during which every AI assistant maker will be deciding whether to pursue Android certification, a decision that shapes the mobile AI landscape for years.
The single thread connecting all of it remains the open-weight offensive. If DeepSeek and Kimi ship on schedule and enterprises start migrating high-volume workloads, the pricing power of closed frontier models erodes in a way that is very hard to reverse. That is the story to watch this week and probably for the rest of the quarter, and we will be tracking it daily. Grok 4.5's aggressive pricing already showed how fast the floor can move, as our Grok 4.5 hands-on review documented.
Frequently Asked Questions
What did the EU order Google to do about AI?
The European Commission adopted binding Digital Markets Act requirements ordering Google to open Android to rival AI assistants, granting them voice activation and cross-app capabilities across 11 Android feature groups subject to certification and user consent, and to share anonymized search ranking, query, click, and view data with competitors. Search data sharing begins January 2027 and Android interoperability is due by July 2027.
Why is Gemini 3.5 Pro delayed again?
Gemini 3.5 Pro reportedly missed its July 17 target for the third time after falling short on coding and complex reasoning in testing, following Google scrapping the original base model in June and restarting pretraining. Google is reportedly exploring a stopgap Gemini 3.6 Flash release, and has published no official model card, pricing, or benchmarks.
Why is Oracle cutting 30,000 jobs?
Oracle is cutting up to 30,000 employees, about 18 percent of its workforce, to free an estimated $8 to $10 billion in annual cash flow to fund AI data center construction for Stargate, the $500 billion infrastructure initiative with OpenAI and SoftBank. Oracle signed a $300 billion five-year cloud contract with OpenAI covering 4.5 gigawatts of capacity.
What is Microsoft's Project Perception?
Project Perception is Microsoft's AI cybersecurity platform that finds and fixes software vulnerabilities using models from Microsoft, OpenAI, and Anthropic together. An orchestration layer routes each task to the best-fit model, using cheap models for triage and frontier models only for complex reasoning, which cuts costs enough to make continuous scanning practical. It is positioned against Anthropic's Mythos-class security offering.
Why did SAP buy Prior Labs?
SAP completed its acquisition of Prior Labs, the pioneer of tabular foundation models, and committed over 1 billion euros across four years to build it into a European frontier AI lab. SAP concluded the biggest untapped enterprise AI opportunity was not language models but AI purpose-built for the structured data in business databases. Prior Labs' TabPFN model series was published in Nature.
When do Kimi K3's open weights release?
Moonshot AI has promised Kimi K3's open weights by July 27, 2026, about eleven days after the July 16 API launch. Combined with DeepSeek V4's stable release on July 24, the final week of July is the largest concentration of open-weight releases the industry has seen.
Recommended Blogs
ā AI News Today July 18 2026: 18 Biggest Stories
ā AI News Today July 17 2026: 20 Biggest Stories
ā AI News Today July 16 2026: 15 Biggest Stories
ā Best AI Models July 2026: Full Ranked Leaderboard
ā Grok 4.5 Review: xAI's Coding Model Tested
ā AI Coding Tools 2026: The Complete Hub
Resources & Community
Join our community of 70,000+ AI enthusiasts and learn to build powerful AI applications! Whether you're a beginner or an experienced developer, Build Fast with AI helps you understand and implement AI in your projects.
ā Website - buildfastwithai.com
ā LinkedIn - Build Fast with AI
ā Instagram - @buildfastwithai
ā Founder Twitter - @satvikps
ā Twitter - @BuildFastWithAI
Agentic AI Launchpad 2026
A structured 6-week cohort program that takes you from AI basics to building and deploying real-world agentic AI systems. Includes live sessions, expert mentorship, project reviews, and a builder community network.
Ready to go from learning to building? Join the next cohort ā Agentic AI Launchpad 2026
Free AI Resources
Access free tools, workshops, and micro-learning to keep building:
ā AI Workshops - Free resources, upcoming events & past recordings
ā Unrot - Learn AI in 5 minutes a day (free micro-learning app)
This week brings DeepSeek V4 on July 24 and free Kimi K3 weights on July 27. Follow Build Fast with AI and subscribe so each recap lands before your standup.
References
ā Computerworld - Google Must Open Android to Rival AI Agents, EU Orders
ā US News - EU Forces Google to Share Search Data and Open Android
ā TechRepublic - Microsoft's Project Perception Could Challenge Anthropic's Mythos
ā SAP News - SAP Completes Prior Labs Acquisition
ā Tech.eu - SAP Acquires Prior Labs in a Billion-Euro Deal
ā Capacity - Oracle Cuts Up to 30,000 Jobs to Fund AI Data Centre Push
ā VentureBeat - Moonshot AI Releases Kimi K3, Largest Open Model Ever
