Today's brief — CIO Daily Brief: signal-ranked tech news, podcasts & intelligence for CIOs and IT leaders

Perplexity launches Portable Computer, a local AI agent platform running fully on-device with zero token costs, starting with Nvidia DGX Spark and RTX Linux PCs (Michael Nuñez/VentureBeat)

Perplexity's new Portable Computer platform enables fully on-device AI agent execution with zero API token costs, eliminating reliance on cloud inference and reducing operational expenses while improving data privacy and latency for enterprises. This on-premise deployment model starting with Nvidia hardware represents a significant shift toward self-sovereign AI infrastructure, requiring IT organizations to reconsider their AI architecture strategy and infrastructure investments. The move signals a broader market trend toward edge AI deployment that could fundamentally alter cloud AI economics and necessitate new skills, procurement processes, and governance frameworks within IT organizations.

Waymo plans to offer ride-hailing services in Germany, starting with Munich, by the end of 2027, marking its third international market and the first in the EU (Natalie Lung/Bloomberg)

Waymo's expansion into Germany by end of 2027 signals the accelerating commercialization of autonomous vehicle technology in regulated EU markets, creating new competitive dynamics in mobility services and indicating that autonomous systems are moving from pilot programs to mainstream deployment. This expansion has significant implications for IT organizations supporting transportation, logistics, and smart city infrastructure, as it will drive demand for edge computing, real-time data processing, and cybersecurity capabilities required to operate autonomous fleets at scale. Technology leaders should prepare for increased investment requirements in infrastructure modernization and consider how autonomous vehicle ecosystems will integrate with existing enterprise systems and urban technology platforms.

OpenAI says its Jalapeño chip delivered 1.5x-1.9x more AI work per watt and 1.7x-3.6x lower latency vs. Nvidia chips across GPT-OSS, DeepSeek R1, Kimi K2.5 1T (Emma Roth/The Verge)

OpenAI's new Jalapeño chip demonstrates significant performance advantages over Nvidia's current offerings, delivering 1.5-1.9x better energy efficiency and 1.7-3.6x lower latency across multiple AI models, which could substantially reduce operational costs and improve application responsiveness for AI-intensive workloads. This development signals a fundamental shift in the AI infrastructure landscape, with major AI companies vertically integrating custom silicon to reduce dependency on Nvidia and potentially reshape vendor relationships and IT procurement strategies. CIOs should anticipate increased competition in the GPU market, potential cost pressures on Nvidia alternatives, and new architectural decisions around chip selection for AI deployments.

OpenAI’s Jalapeño chip is built for fast inference at scale, benchmarks show

OpenAI's Jalapeño inference chip demonstrates significant performance advantages over current state-of-the-art processors (including Nvidia Blackwell), delivering higher token throughput per watt and lower latency—critical metrics for cost-effectively scaling AI workloads at enterprise scale. Deployment will begin in limited volumes by end of 2026 with broader availability in 2027, signaling a shift toward proprietary, full-stack AI infrastructure that could reshape cloud economics and competitive positioning in the AI services market. CIOs should anticipate potential shifts in AI infrastructure costs and vendor strategies as custom silicon becomes increasingly central to AI deployment efficiency.

Apple debuts its ‘most powerful chip ever’ in M5 Ultra and M6

Apple's new M5 Ultra and M6 chips represent a significant shift toward on-device AI processing, with the M5 Ultra delivering unprecedented computational power (36-core CPU, 80-core GPU) for enterprise AI workloads while the M6 enables local LLM fine-tuning on consumer Macs with 30% improved AI performance. This architecture positions Apple as a privacy-focused alternative to cloud-based AI solutions, compelling IT organizations to evaluate Apple Silicon capabilities for AI-intensive workflows and consider the competitive implications of native on-device processing for enterprise security and compliance strategies. Organizations relying on traditional cloud-based AI infrastructure should assess whether Apple's secure, on-device approach creates new opportunities or threats to their existing technology roadmaps.

OpenAI says its Jalapeño chip can power faster AI responses than the competition

OpenAI's new Jalapeño chip delivers 1.5-1.9x better energy efficiency and 1.7-3.6x lower latency compared to Nvidia's superchips for AI inference workloads, fundamentally shifting the economics of AI infrastructure deployment. With volume production ramping through 2027, this custom silicon represents a strategic move by OpenAI to reduce dependency on external chip suppliers while enabling faster, more cost-effective AI service delivery. IT leaders must prepare for a fragmented AI accelerator landscape where multiple vendors offer specialized chips, requiring updated procurement strategies and infrastructure planning to optimize for both performance and total cost of ownership.

Apple says M5 Ultra uses a quad-die 3nm architecture, "a first for Apple silicon", with inter-die bandwidth of 4.4TB/s+ and up to a 36-core CPU and 80-core GPU (Steve Dent/Engadget)

Apple's new M5 Ultra chip introduces a groundbreaking quad-die 3nm architecture with exceptional inter-die bandwidth (4.4TB/s+) and scaling to 36-core CPU and 80-core GPU, positioning Mac Studio as a serious alternative to traditional high-performance workstations for compute-intensive tasks. This advancement has significant implications for IT organizations managing creative and engineering workflows, potentially reducing total cost of ownership while improving performance for video rendering, 3D modeling, and scientific computing. The continued evolution of Apple silicon with the M6's 2nm process technology signals Apple's commitment to vertical integration and performance leadership, requiring CIOs to reassess hardware procurement strategies and Mac adoption in performance-critical departments.

Apple unveils M6, a 2nm chip with a 12-core CPU and GPU, and up to 32GB of unified memory, saying it provides "the world's fastest single-threaded performance" (Jay Peters/The Verge)

Apple's new M6 chip (2nm process with 12-core CPU/GPU and up to 32GB unified memory) represents a significant performance leap in computing power, potentially reshaping IT infrastructure decisions around macOS environments and edge computing capabilities. For technology leaders, this advancement may impact hardware refresh cycles, developer productivity investments, and strategic decisions around Mac adoption in performance-critical workloads. Organizations should evaluate whether M-series performance gains justify migration investments or influence upcoming device procurement strategies.

Apple unveils a Mac mini with M6 and M5 Pro, with up to 4x faster AI performance and up to 2x faster graphics, for $899+ with M6 and $1,699+ with M5 Pro (Apple)

Apple's new Mac mini M6 and M5 Pro models deliver significant performance improvements—up to 4x faster AI performance and 2x faster graphics—at competitive price points ($899+ and $1,699+ respectively), making advanced AI and machine learning capabilities more accessible for enterprise workloads. For IT organizations, this represents an opportunity to evaluate cost-effective alternatives for AI-intensive development, content creation, and data processing tasks while maintaining Apple's ecosystem advantages. The enhanced performance-to-price ratio could influence refresh cycles and deployment strategies, particularly for teams requiring on-device AI processing and reduced cloud computing dependencies.

Apple updates the Mac Studio with M5 Max and M5 Ultra, with up to 4.3x faster AI performance, faster graphics, and up to 512GB of unified memory for $2,499+ (Apple)

Apple's updated Mac Studio with M5 Max/Ultra processors delivers 4.3x faster AI performance and up to 512GB unified memory, positioning macOS as a competitive platform for enterprise AI workloads and data-intensive applications without reliance on cloud infrastructure. For IT organizations, this expansion of local AI capabilities at the $2,499+ price point creates new opportunities for on-premises ML deployment, enhanced data privacy, and potential cost optimization compared to cloud-based alternatives. The unified memory architecture and graphics improvements also strengthen Apple's enterprise appeal for creative workflows and engineering teams, warranting reassessment of Mac infrastructure in AI-forward IT strategies.

The state of AI in 2026: On the road to ROI

Organizations are transitioning from AI experimentation to measurable return on investment, with enterprise adoption accelerating across operational and customer-facing use cases. IT leaders must shift focus from pilot programs to scaling proven AI solutions while addressing critical challenges around data quality, talent retention, and responsible AI governance. Success in 2026 will depend on aligning AI investments with specific business outcomes and establishing clear accountability for AI project ROI.

Apple rescues Hide My Email feature from the privacy scrap heap

Apple reversed a planned change to its Hide My Email privacy feature that would have made masked email addresses easily identifiable to websites, responding to user backlash that the modification would undermine the service's core privacy value proposition. This decision demonstrates the growing importance of user privacy expectations and the reputational risk of weakening privacy protections, even among premium subscribers. For IT organizations, this underscores that privacy features must maintain genuine technical protections rather than providing only superficial anonymity, as any perceived weakening can erode user trust and compliance posture.

Autonomous trucking startup Gatik raised $200M, two months after striking a multi-year commercial agreement with PepsiCo, bringing its total funding to ~$500M (Kirsten Korosec/TechCrunch)

Autonomous trucking startup Gatik's $200M funding round and multi-year commercial agreement with PepsiCo signals that autonomous logistics technology is moving from pilot phase to production-scale deployment, creating both operational efficiency opportunities and supply chain disruption risks for enterprise IT leaders. This trend indicates that autonomous vehicle integration will soon require IT organizations to modernize their fleet management, data infrastructure, and cybersecurity capabilities to support AI-driven logistics systems. Technology leaders should begin evaluating autonomous vehicle technologies and their integration requirements now, as first-mover enterprises will gain competitive advantages in supply chain optimization and cost reduction.

Self-driving truck startup Gatik raises $200M following PepsiCo deal

Autonomous vehicle startup Gatik has raised $200M to scale its proven middle-mile delivery operations, which now include $600M in contracted commercial revenue with major customers like PepsiCo, Kroger, and Walmart. This signals that autonomous vehicle technology has matured beyond pilot phases into reliable commercial operations, creating potential disruption across supply chain and logistics operations that IT leaders must monitor. For enterprise IT organizations, this validates the business case for autonomous logistics infrastructure and signals the need to plan for integration of driverless fleet management systems into existing supply chain and transportation technology stacks.

Apple’s latest Mac Mini runs on a new M6 chip, and starts at $899

Apple's new M6 Mac Mini ($899 base model) delivers 4x better AI performance and doubled storage/graphics speed compared to its M4 predecessor, positioning it as an increasingly attractive platform for enterprise AI workloads and edge computing deployments. The stronger performance-to-price ratio, coupled with growing demand for local AI agent execution, signals a shift in how organizations can cost-effectively deploy on-device AI capabilities without reliance on cloud infrastructure. IT leaders should evaluate whether Mac Mini's improved specifications and competitive pricing create opportunities to expand macOS-based development environments and reduce cloud compute expenses for AI-adjacent workloads.

Apple raises Mac mini and Studio prices: the M6 and M5 Pro Mac mini are $899+ and $1,699+, $100 above M4 models, and the M5 Ultra Mac Studio is $5,499+, up $200 (Nicole Nguyen/Wall Street Journal)

Apple has increased Mac mini and Studio pricing by $100-$200 across the M5/M6 lineup, positioning premium development and creative workstations at higher price points amid AI infrastructure demand and semiconductor constraints. This pricing strategy reflects Apple's confidence in performance upgrades while signaling potential supply chain pressures that could affect IT procurement timelines and budget planning. Technology leaders should expect continued price pressure on high-performance computing equipment and may need to accelerate purchasing decisions or explore alternative platforms if stock availability becomes constrained.

WhatsApp upgrades its two-step verification, letting users replace the six-digit PIN with a longer alphanumeric password, and adds support for multiple passkeys (Aisha Malik/TechCrunch)

WhatsApp is enhancing its authentication security by replacing limited six-digit PINs with longer alphanumeric passwords and enabling multiple passkey support, significantly reducing account compromise risks for enterprise users. This upgrade aligns with modern zero-trust security frameworks and passwordless authentication trends, requiring IT organizations to update security policies and user training programs to accommodate the new verification methods. For CIOs managing enterprise messaging security, this represents an opportunity to strengthen endpoint protection while reducing password-related help desk burden through passkey-based authentication.

MySQL CDC to BigQuery: what periodic syncs miss, and how binlog avoids it

MySQL Change Data Capture (CDC) using binary logs provides superior data integrity compared to traditional periodic batch syncs, capturing all deletes and intermediate updates in order rather than relying on snapshot comparisons that miss critical changes and strain production databases. Organizations must configure specific MySQL prerequisites—including ROW format binlog with FULL images, proper replication privileges, and adequate binlog retention—to ensure reliable data pipelines to BigQuery, with managed platforms reducing implementation complexity. The strategic shift from "how do we make this faster" to "what are we currently unable to see" is critical for data-driven decision-making, as incomplete data arriving on schedule creates greater business risk than occasional latency with guaranteed accuracy.

Apple's new Mac mini, featuring M6 and M5 Pro

Apple's new Mac mini with M6 and M5 Pro chips delivers up to 4x faster AI performance and 2x faster graphics, positioning it as a viable platform for on-device AI workloads and enterprise agentic computing while maintaining a compact form factor. For IT organizations, this represents a strategic shift toward supporting edge AI inference on desktops, requiring updated procurement strategies, IT infrastructure planning for higher-performance endpoints, and consideration of Apple Silicon in enterprise deployments. The enhanced connectivity (Wi-Fi 7, optional 10Gb Ethernet) and support for larger AI models (up to 64GB memory on M5 Pro) create both opportunities for AI-driven productivity gains and new management and security considerations for CIOs.

Apple Introduces New Mac Studio with M5 Max and M5 Ultra

Apple's new Mac Studio with M5 Max and Ultra chips delivers up to 4.3x faster AI performance and 512GB of unified memory, enabling enterprises to run frontier-class AI models locally with enhanced privacy and reduced cloud costs. For IT organizations, this positions Mac as a viable platform for AI workloads, data science, and creative workflows, while the Thunderbolt 5 clustering capability (3x performance gain) creates new opportunities for distributed computing without cloud dependencies. CIOs should evaluate Mac Studio's on-device AI capabilities as part of their broader AI strategy, particularly for sensitive workloads requiring data sovereignty and organizations seeking to reduce third-party AI service dependencies.

The New Economics of AI | Martin Casado & Steven Sinofsky

a16z General Partners Martin Casado and Erik Torenberg are joined by Board Partner Steven Sinofsky to explore what recent breakthroughs in AI and mathematics tell us about where the technology is headed, and whether some of the basic assumptions that have governed computing for decades are starting to break. Martin and Steven debate whether AI's progress in mathematics represents a genuine leap in reasoning or simply a new tool for solving problems at a higher level of abstraction. From the four

AI Proficiency: From Users to Builders

As AI continues to reshape how organizations work, companies are increasingly asking what AI proficiency should look like across their workforce, and how they can help employees adapt without simply mandating AI adoption. Our returning guest Mike Lewis, Chief AI Architect at TiER1 Performance, joins Dan and Chris to explore AI proficiency through the L0–L3 framework, with a particular focus on the role of non-technical builders. They discuss AI resistance, identifying the right people to build A

The Gap Between AI Spending and AI Value

It is widely reported that a gap has emerged between enterprise spending on AI and the durable value captured from that spend. Individual employees have enthusiastically adopted coding assistants and chatbots, yet those gains do not seem to be transforming businesses at an organizational level. One of the most important questions in the tech industry today is understanding why AI is not yet delivering returns that match the investment, and what separates the small number of enterprises succeedin

The odds were classified.

Polymarket traders win big on U.S. military insider information. Slovakia deactivates speed cameras with Russian backdoors. TikTok pays $400 million to settle kids' privacy allegations. Hackers infect Android-based car systems with botnet malware. CISA orders quick patching of an actively exploited Zimbra Collaboration Suite vulnerability. SynkLoader malware is built for stealthy access to corporate networks. Dutch authorities fine Uber over $900 million over automated hiring practices. An ATM j

The AI Model Tier List

A viral AI model tier list reveals how much harder it has become to name the “best” model. This episode breaks down where today’s leading models belong, why cost and speed increasingly matter alongside intelligence, and how businesses are assembling model stacks that combine premium and open models. In the headlines: Hugging Face explores a sale, NVIDIA expands its open-model ambitions, and Dr. Dre embraces AI music.Executive Agent Leadership - Returns in September -- Learn how to use agents - ⁠

Why Medical AI Needs a Referee | Protege's Engy Ziedan

Daisy Wolf and Eva Steinman are joined by Engy Ziedan, co-founder and Chief Scientific Officer of Protege, to discuss why medical AI has a measurement problem, and why scoring well on a benchmark doesn't necessarily mean a model is ready for the hospital. Engy explains why healthcare AI needs independent evaluations that go beyond static exams and measure how models actually perform in real-world clinical workflows. They explore the risks of subtle bias and misalignment, why the same model can r

Simplify First, Scale AI Next with Vulcan Materials CIO Krzysztof Soltan

Better AI starts with better data, and for Vulcan Materials CIO Krzysztof Soltan that means simplifying first. In this episode of Technovation, Krzysztof Soltan joins Peter High to discuss how Vulcan Materials is modernizing its technology foundation across more than 400 aggregate operations. He explains why reducing system complexity, strengthening master data governance, and investing in talent are foundational to the company’s AI ambitions. The conversation also explores Vulcan’s multi-year d

NVIDIA's Pivot from Chipmaker to Financier

The Real Future of AI and Work

AI’s impact on work goes far beyond job losses. Drawing on Every’s new Thesis Statements project, NLW explores how AI could transform what individuals do, how companies operate, which skills become valuable and what becomes possible when intelligence is abundant.Thesis Statements: https://every.to/thesis-statementsExecutive Agent Leadership - Returns in September -- Learn how to use agents - ⁠https://training.besuper.ai/⁠Free Webinar - Agentic Loops for Knowledge Workers - 8/26/26 26pm ⁠https://

Building a secure space internet. [T-Minus: Space-Cyber Briefing]

As space infrastructure has continued to expand, developing secure space systems has become just as important as launching the spacecraft themselves. In this week's episode, host Maria Varmazis sits down with Filip Rezabek, co-founder and CTO of Space Computer, to talk about some of the technologies being created to secure space infrastructure in orbit. The two discuss the importance of establishing a chain of trust in space and the challenges of securing hardware against supply chain attacks. L

Martin Casado on Where the Value Is Going in AI

Martin Casado joins MTS hosts Theo Jaffee and Sophia Dew to unpack where value is actually accruing in AI, why this technology cycle looks fundamentally different from previous waves, and whether the frontier labs will ultimately capture most of the market. Martin explains why AI has turned venture into a scale-up capital game, where small teams can productively deploy extraordinary amounts of money, and why the relationship between capital, innovation, and growth has never been tighter. He lays

A RAT in the spreadsheet. [Research Saturday]

Today we are joined by Aaron Beardslee, Manager of Threat Research at Securonix, discussing "Analyzing SHEET#CREEP: SHEETCREEP is up again with different config obfuscation." Securonix researchers have identified an evolved version of the SHEETCREEP espionage campaign, using a diplomatic-themed ISO phishing lure to deliver a C# remote access trojan targeting Indian diplomatic interests. The malware abuses the Google Sheets API as a stealthy command-and-control channel, with researchers identifyi

Why Everyone Suddenly Hates AI Data Centers

Opposition to AI data centers has become one of America’s fastest-moving bipartisan political issues, driven by concerns about electricity, water, noise, property values, jobs, and a deeper mistrust of Big Tech. NLW provides an end-to-end primer on what data centers actually do, why public sentiment has turned so sharply, where the loudest claims fall apart, and how transparency, community control, and direct local benefits could offer a path forward.Executive Agent Leadership - Returns in Septe

The guest nobody invited.

CISA orders patching of TrueConf Server vulnerabilities. LockBit threatens release of stolen banking data. Researchers disclose a critical type confusion vulnerability in a Node.js library. A new Agent Tesla v4 campaign introduces enhanced evasion techniques. A novel malware delivery technique abuses FTP server banners to hide commands. Apple patches a critical image-processing flaw. A North Korean software supply chain attack targets the Rust ecosystem. Latvian officials resign following a majo

How Microsoft Is Securing the Agentic Enterprise | Aaron Zollman

a16z's Joel De La Garza is joined by Aaron Zollman, Deputy CISO at Microsoft Gaming, to discuss how security teams can embrace AI agents without losing control. Aaron shares Microsoft's experience with OpenClaw, from the initial instinct to ban it to figuring out how to make it safe to use. They unpack what agents mean for identity, permissions, containerization, and monitoring, as well as how AI is shifting the CISO's role from saying "no" to safely enabling new technology. They also explore wh

9 AI Techniques You Probably Haven't Tried

Even experienced AI users can fall behind as new features and working methods arrive. NLW breaks down nine techniques worth trying now—from live voice mode, workflow teaching, custom skills, and Claude’s /design command to team agents, GrokBot, local models, and deceptively useful two-word prompts. In the headlines: an AI-assisted personalized cancer vaccine clears a Phase III trial; OpenAI introduces private safety processing; and Replit launches Free Mode with GPT-5.6 Luna.Free Webinar - Agent

The robots have gone bananas.

Federal agencies warn of an active campaign targeting critical infrastructure. Citrix races to patch critical NetScaler flaws. More than 50,000 exposed Stripe API keys raise fraud concerns. Black Hat and DEF CON attendees are targeted in a new social engineering campaign. Atlassian, Splunk, and Cisco fix hundreds of vulnerabilities. A new Android banking trojan adds an unusual twist. A healthcare breach impacts 3.8 million people. SilkParasite expands cyberespionage in Central Asia. And CISA eye

How Global Networks Are Reshaping Startup Success

Elena Burger is joined by a16z’s Angela Strange and Gabriel Vasquez to discuss the rise of the "borderless founder": entrepreneurs who bring the networks and insights of their home markets together with the talent, capital, and speed of Silicon Valley to build global companies. Angela and Gabriel trace how a16z's international investing efforts grew from early work in Latin America into a broader global network, and why AI has accelerated the flow of founders and talent between Silicon Valley an

Aetna’s Nathan Frank on Scaling AI to Give Healthcare Workers Time Back

What happens when AI gives healthcare workers more time to do the work only humans can do? Nathan Frank, Chief Digital and Technology Officer of Aetna, joins Peter High to discuss how Aetna is scaling AI across healthcare while keeping people at the center. Frank shares how AI-enabled workflows are giving care-management nurses 90 minutes back per day, reducing the administrative burden of navigating clinical notes and multiple systems. He also explains how Aetna is building on years of investme

AI and the New Global Security Landscape

The conversation about AI often focuses on software, automation, and the race between attackers and defenders in code. However, some of the most consequential risks lie further afield, in domains where a mistake is measured in human lives. Advanced models can now offer step-by-step guidance toward chemical and biological weapons, and militaries are already folding AI into targeting and battlefield assessments. These are no longer speculative fears confined to the AI doomer crowd. They are docume