BTW.Media
Daily editionStrategic Internet Intelligence
Sign InRegister

Briefing Desk

Latest Briefings

Concise reporting on the developments shaping internet governance and infrastructure. Browse each area for recent news, context and watchpoints.

Coverage

Market / Trends / North America Trends / North America Cloud Services Trends

In this section: 2 briefings

  1. AWS’s Lusby withdrawal ends one site-plan process, not the data-centre argument

    Calvert County says Amazon Data Services, also referred to as AWS, formally withdrew the Conceptual Site Development Plan for the Calvert Technology Center on 4 August. The application had passed acceptance and a first review but had not reached construction. Local reporting describes an approximately 500MW, eight-building concept near Calvert Cliffs. AWS gave no public reason. The precise state change is therefore narrow but consequential: one large proposal has left the county’s review queue, while the underlying questions about land, power, water and future applications remain open.

  2. Zoom’s status object kept Meetings operational while reporting degraded quality

    Zoom says a subset of users in its US region experienced degraded meeting quality for ten minutes on 2 August. The incident record appeared about 48 minutes after that interval ended, pairing an operational component state with a narrative account of impaired experience.

Coverage

Market / Trends / Global Trends / Global Cloud Services Trends

In this section: 16 briefings

  1. Cloudflare’s Workers regressions exposed two kinds of release contract

    Cloudflare closed two separate Workers incidents on 4 August: a runtime exposed an unintended `Temporal` global whose clock reported 1970, while a deployment assertion rejected Workers using `nodejs_compat` with a new compatibility date. Neither was described as a platform-wide outage, and Cloudflare did not give them a common cause. Together, however, they show why a release contract must cover both what an application detects at runtime and what the control plane permits at deployment.

  2. Palo Alto Networks moves Unit 42 intelligence into Cortex—and closer to the customer’s switching cost

    Palo Alto Networks launched Unit 42 Threat Intelligence on 3 August through an embedded Cortex product and a separate analyst-led service. The offer is designed to sell something more valuable than another threat feed: a judgement about which threats matter inside a customer’s own environment, and a path from that judgement into security operations. That can remove work for buyers already committed to Cortex. It can also make the platform harder to leave.

  3. Twilio’s Deepgram model outage showed why a backup is not a failover plan

    For two hours and 43 minutes on 3 August, Twilio tracked a major incident in which ConversationRelay customers configured with Deepgram’s nova-2 or nova-3 speech models were described as “hard down.” Other named model paths remained available, but Twilio’s own advice contained the crucial qualification: a customer should consider Google only if that option had already been validated for the use case.

  4. DigitalOcean renewed internal certificates after a 234-minute control-panel access incident

    For nearly four hours on 3 August, some DigitalOcean customers intermittently could not enter the Cloud Control Panel. The visible clue was an “mTLS verification failed” message; the disclosed corrective act was renewal of affected internal certificates. That is enough to locate the incident in an administrative trust path, but not enough to call it a cloud-wide outage, a certificate expiry or a security breach.

  5. Cloudflare’s Workers Builds incident shows why software can be live but unable to change

    Cloudflare marked an August 3 Workers Builds incident resolved after one hour and 51 minutes. The revealing moment came earlier: at 15:38 UTC, builds were no longer failing, yet users could still encounter delays. That interval separates two kinds of availability—keeping an existing deployment in place and retaining the ability to compile and release the next one. Cloudflare did not disclose a root cause, customer count or runtime impact, so the event supports a precise lesson about change capacity, not a claim of a global Workers outage.

  6. GitHub restored Copilot after multi-model request failures exposed a shared work queue

    GitHub resolved a 92-minute Copilot incident after chat and agent requests across multiple models suffered degraded availability. The service recovered, but the public record leaves enterprises without the denominator needed to judge how much work failed, waited or succeeded on retry.

  7. NTT's OptQC pact adds capital to a roadmap, not a million-qubit machine

    NTT and OptQC have turned a research relationship into a capital and business alliance, with NTT planning to invest in the University of Tokyo spin-out. The agreement creates a longer operating horizon for photonic quantum computing, but the price, stake and funding date remain undisclosed, while the flagship machine is still a 2030 target.

  8. Twilio sees Safaricom Kenya receipt recovery after 57 minutes, but the cause stays undisclosed

    Twilio says SMS delivery receipts on a route to Safaricom subscribers in Kenya recovered after a 57-minute public incident window. The underlying messages may have succeeded while their acknowledgements arrived late. Service is being monitored, not declared resolved, and the operator has not explained the cause or quantified the affected traffic.

  9. Zoom disclosed a 76-minute Marketplace and AI Companion degradation only after it had ended

    Zoom says users experienced degraded service across Marketplace and AI Companion from 00:00 to 01:16 UTC on 2 August. The incident record appeared roughly 29 minutes later, naming two service families but no affected function, cause, scale or geography.

  10. Cloudflare’s 15-minute Istanbul network incident left a precise window but almost no diagnosis

    Cloudflare says elevated latency and connection errors affected Istanbul for 15 minutes on 1 August. The unusually exact interval is the firmest part of a record that identifies no product, route, cause or scale—and whose only visible notice appeared after the reported impact had ended.

  11. GitHub restored GPT-5.6 Luna after 74 minutes, but the incident exposes a model-routing dependency

    GPT-5.6 Luna became degraded across GitHub Copilot products and IDE surfaces because of an unnamed upstream model provider on 1 August. GitHub offered a practical escape route—choose another model or Auto—but published no measure of how many requests failed or what functional trade-off the substitution imposed.

  12. Twilio’s Globe Philippines incident delayed the receipt signal, not necessarily the SMS

    For 2 hours and 42 minutes on 1 August, Twilio reported delayed SMS delivery receipts from its phone numbers to Globe Telecom subscribers in the Philippines. Messages could still succeed. That distinction matters because applications often treat the receipt as a control signal even when the underlying transport outcome is different.

  13. Cloudflare’s RealtimeKit join failures lasted 65 minutes, while its incident metadata says otherwise

    Cloudflare says some RealtimeKit users faced slow socket connections and failed meeting joins from 13:10 to 14:15 UTC on 1 August. Yet the incident object records both creation and resolution at 13:10:30. The narrative is operationally useful; the metadata cannot describe the same impact window.

  14. GitHub took 17 minutes to name Fable 5 after upstream model errors began

    GitHub’s second public Copilot model-provider incident on 1 August began with generic reports of upstream errors and only later identified Fable 5. Mitigation followed 28 seconds after that model-specific disclosure. The chronology shows fast recovery once the route was named, but not whether discovery itself or public attribution consumed the preceding 17 minutes.

  15. Cloudflare’s IAD 5XX episode lasted 256 minutes but was disclosed in one retrospective update

    Cloudflare says an increased level of HTTP 5XX errors affected Ashburn, US (IAD), from 18:45 to 23:01 UTC on 31 July. Yet the public incident contains only one visible update, timestamped more than two hours after that interval ended. The fault is closed; the operational sequence, affected product and scale remain largely invisible.

  16. Twilio’s Claro Colombia short-code delay was still under investigation at the 20-minute cutoff

    Twilio opened a minor incident at 02:57:20.918 UTC on 1 August for SMS delivery delays from a subset of its short codes to Claro subscribers in Colombia. At the fixed 03:17:59 cutoff, 20 minutes and 38.082 seconds had elapsed. There was no identified cause, recovery signal or resolution to report.

Coverage

Market / Trends / Global Trends / Global National Telecom Trends

In this section: 2 briefings

  1. HPE’s Telco Solutions sale moves the people and product decisions—not just the software

    HCLTech completed its acquisition of HPE’s Telco Solutions business on 4 August, bringing nearly 1,400 engineering and telecom specialists in 39 countries into the buyer’s operating model alongside intellectual property, OSS, HSS and 5G subscriber-data capabilities. The closing is a transfer of control over product roadmaps, support and specialist labour. Customers now need evidence that those operating promises survive the change of owner; the announcement’s broad AI and network-modernisation claims do not provide it yet.

  2. AccelerComm completes Eclipse Space physical-layer design

    AccelerComm has completed physical-layer design work for Eclipse Space’s planned direct-to-device and broadband satellite platform.

Coverage

Market / Trends / Asia-Pacific Trends / Asia-Pacific Cloud Services Trends

In this section: 1 briefing

  1. LG Uplus buys an MDR operating loop—but leaves the ownership perimeter undisclosed

    LG Uplus announced on 4 August that it had acquired PAGO Networks, a managed detection and response company whose DeepACT platform joins technology with round-the-clock analyst operations. The strategic shift is from selling security components toward controlling detection, judgement and response as a continuous service. Yet the price and acquired stake were not disclosed. Before customers treat the deal as an integrated security platform, LG Uplus must define what it owns, which teams hold response authority and how telemetry and contracts will move.

Coverage

Market / Trends / Asia-Pacific Trends / Asia-Pacific National Telecom Trends

In this section: 1 briefing

  1. China Mobile’s True filing turns an exit story into a possible portfolio trim

    True Corporation used a 4 August filing to reject the claim that China Mobile International intended to sell its entire 7.81% holding. According to True’s CFO, the shareholder was only assessing a possible sale of a small portion and intended to remain a long-term shareholder and strategic partner. The filing changes the information state, not the ownership state: no size, buyer, price, agreement or transfer was disclosed. That distinction matters more than the headline estimate of what the whole stake might be worth.

Coverage

Market / Trends / Europe and Middle East Trends / Europe and Middle East Datacenter Trends

In this section: 1 briefing

  1. Start Campus targets 2027 opening for 200MW SIN02

    Start Campus plans to open its 200MW SIN02 data centre in 2027, alongside Nscale’s planned deployment of more than 66,000 Rubin GPUs.

Coverage

Market / Trends / North America Trends / North America National Telecom Trends

In this section: 4 briefings

  1. NTIA sends 4.4GHz spectrum study plan to Congress

    NTIA has sent the 4.4GHz study plan to Congress, starting a 60-day review before agencies can receive funding for engineering work.

  2. TELUS cuts dividend 55% as digital unit loses value

    TELUS is retaining more cash for debt reduction as it focuses resources on connectivity and reviews assets outside its core telecom business.

  3. Hughes has put part of its satellite business behind a court wall, not switched it off

    Hughes Satellite Systems Corporation and eleven wholly owned subsidiaries entered Chapter 11 on 2 August. The petitions move debt enforcement and major restructuring decisions into a supervised legal process while management says customer operations should continue. The useful question is therefore not whether “Hughes went bankrupt” in the abstract. It is which companies are inside the case, who controls the next decisions, what cash and contracts keep services running, and which claims still need court or operational proof.

  4. Michigan’s nine water-system reports widen the cyber incident without showing unsafe water

    Michigan says nine water systems reported malicious activity consistent with a federal alert. The systems kept operating safely and local teams addressed the issues, leaving a serious operational-technology investigation but no disclosed public-health impact or named culprit.

Coverage

Market / Trends / North America Trends / North America Datacenter Trends

In this section: 13 briefings

  1. Dozens protest Meta’s planned Alberta data centre

    Local opposition has emerged after Meta secured key permits for its 1GW Alberta data centre, focusing attention on grid use, water and project oversight.

  2. Apple shares fall on chip shortage warning

    Apple’s weaker outlook shows how AI-driven memory demand and separate processor shortages are raising costs and limiting device supply.

  3. Iowa’s data-centre tool turns local controversy into questions that can be audited

    The University of Iowa has released a decision tool for counties, cities and communities considering data-centre development. It does not tell officials to approve or refuse a project. Its value is more procedural: it converts arguments about water, power, taxes, land and construction into questions that can be answered, recorded and—where a deal proceeds—made enforceable.

  4. Texas puts a proof gate in front of its 474GW data-centre queue

    Texas has ordered PUCT and ERCOT to audit every data-centre project advancing through the grid connection process before it can move forward. The state says requests now exceed 474GW—more than five times ERCOT’s record peak demand—and that roughly 90% of new power requests come from data centres. Those figures describe a queue, not future consumption. The practical change is that a project must now make its power, water, subsidy and ownership claims comparable before scarce grid capacity is allocated.

  5. Lakeland’s 50MW line buys a year to write rules before approving concentrated load

    Lakeland’s City Commission voted 4-3 for a twelve-month moratorium on large-scale data centres and other proposed customers with peak demand of at least 50MW. The pause covers the city’s acceptance, processing, review and approval of relevant development applications while officials revisit land use, utility effects and development standards. It is neither a permanent ban nor a rejection of Project Swan. The city is changing the sequence: define the obligations of unusually concentrated load before granting it an administrative path.

  6. Google’s 155MW Oklahoma solar deal has contracted an output, not delivered a data-centre load

    Google has agreed to buy the entire output of RWE’s planned 155MWac Crooked Creek Solar project for 15 years. The contract can give the project a durable buyer and give Google a long-term renewable-energy position in the Southwest Power Pool. It does not mean the plant is operating, 155MW is flowing to a Google facility, or every hour of Google demand is matched by carbon-free generation. Between signature and expected commercial operation in 2028 sit construction, interconnection, production and settlement risks.

  7. Yondr has bought a Manassas site; 72MW still has to climb the entire delivery ladder

    Yondr and investment funds managed by Cerberus have acquired a 40-acre site in Manassas, Virginia, where Yondr plans a 72MW data-centre campus with a 2029 ready-for-service target. The acquisition moves the project past the search for land, but not past the harder conversions that turn land into dependable capacity. The announcement does not disclose price, financing, customers, construction start, utility agreement, permits, building count or energisation milestones. Its importance lies in control of a site in Northern Virginia; its uncertainty lies in every step between that control and an operating rack.

  8. Prysmian’s $3.8 billion Atkore agreement is a route to control, not control itself

    Prysmian has agreed to pay $95 in cash for each Atkore share in a transaction carrying an enterprise value of about $3.8 billion. The industrial case reaches from electrical conduit to data-centre construction, but every claimed benefit still sits beyond a shareholder vote, several regulatory clearances and the legal moment when ownership changes.

  9. Brookfield, NextEra plan $100bn Kentucky data-centre campus

    Brookfield and NextEra plan a 1.8GW Kentucky campus with dedicated gas generation, battery storage and local utility service.

  10. PJM proposes capacity rules for large data centres

    PJM wants new data centres to pair grid connections with firm capacity or accept earlier curtailment during regional shortages.

  11. Lake to River reports surge in data-centre inquiries

    Lake to River reports nearly 10 data-centre leads in eastern Ohio, but none have reached a named project, site selection, or permit filing.

  12. El-Sayed campaigns on binding data-centre rules

    Michigan’s Senate primary is testing voter support for binding conditions on data-centre power, water, jobs and community benefits.

  13. Woodburn council weighs data-centre moratorium

    A moratorium would affect projects only after Woodburn defines which permit actions city staff must stop and from what date.

Coverage

Market / Trends / Global Trends / Global Datacenter Trends

In this section: 1 briefing

  1. Data centre executives challenge ‘energy guzzler’ label

    Industry executives argue that power source, computing output and cooling design give a fuller picture of data-centre resource use.

Coverage

Market / Trends / Europe and Middle East Trends / Europe and Middle East Cloud Services Trends

In this section: 2 briefings

  1. Cloudflare’s London egress incident exposed the narrow risk inside a stable outbound identity

    Cloudflare restored Gateway service after a two-hour incident in which customers using dedicated IPv4 egress addresses homed in London could have been unable to reach the public Internet. The public record is precise about the affected path and the five operational states, but sparse about scale and cause. That combination matters. A dedicated egress address gives an organisation a durable identity for allowlists and policy, yet it also concentrates outbound traffic on a provider-managed dependency whose failure can interrupt destinations that are otherwise healthy.

  2. EU AI transparency duties take effect with separate obligations for providers and deployers

    Article 50 of the EU AI Act now requires in-scope systems to disclose AI interaction, embed detectable machine-readable markings in synthetic output, and label specified biometric, deepfake and public-interest uses. The duties are binding; signing the supporting code remains voluntary.

Coverage

Market / Trends / Africa Trends / Africa National Telecom Trends

In this section: 2 briefings

  1. Helios Towers raises 2026 guidance again

    Helios Towers raised its tenancy, earnings, and cash-flow forecasts after record first-half additions across Africa and the Middle East.

  2. Twilio saw recovery on a subset of SMS routes to Safaricom, but had not closed the incident

    Twilio reported delays and failures from some alphanumeric sender IDs and long codes to Safaricom subscribers in Kenya. It identified an undisclosed cause, observed recovery after 94 minutes and was still monitoring stability at the fixed cutoff.

Coverage

Market / Trends / Europe and Middle East Trends / Europe and Middle East National Telecom Trends

In this section: 3 briefings

  1. EE launches 8Gbps fibre in Guildford and Woking

    EE has begun selling 2.3Gbps and 8Gbps broadband over Openreach’s XGS-PON network in selected Surrey postcodes.

  2. Derbyshire firefighters use O2 Satellite in wildfire

    Derbyshire firefighters used O2 Satellite to exchange updates and access information while responding to a wildfire outside terrestrial mobile coverage.

  3. Iraq uses AI to forecast 135% Karbala traffic surge

    Iraq is using traffic forecasts to prepare Karbala’s mobile networks for a projected 135% surge during the Arbaeen pilgrimage.

Coverage

Market / Trends / Asia-Pacific Trends / Asia-Pacific Datacenter Trends

In this section: 2 briefings

  1. SoftBank unit's Southeast Asia AI data-centre pact allocates roles, not capital

    SB Telecom Singapore, SC ZEUS Data Centers and Robust HPC Group have divided the work needed to pursue AI data centres across Southeast Asia. The memorandum creates a development chain, but names no site, megawatts, power supply, customer, capital budget or delivery date.

  2. Mitsubishi's August target does not yet restore its quake-hit chip output

    Mitsubishi Electric says two semiconductor factories halted after the Kumamoto earthquake should move back to normal operations during August. The timetable narrows the disruption, but it does not disclose current output, lost production or when customers will again receive normal volumes.