CTO
Every story that matters to a CTO, with one sentence on what it concretely means for them. 246 stories, latest first.
November 2026
Google Cloud adds OpenAI models and Codex agents to Model GardenOpenAI models can also be used within Google Cloud, which may reduce vendor fragmentation.Apple, Microsoft and Meta tighten internal AI usage limitsEven large companies are keeping token costs in check, so plan how you will measure return on investment.OpenAI and JPMorgan Chase to deploy Codex and cybersecurity modelsA large-bank deployment can offer a template for using Codex in a regulated environment.OpenAI to acquire Pierre Computer to strengthen CodexReassess vendor risk in tool choice, because code hosting and the agent market are converging.OpenAI builds new Git backend Pando after Codex agents drive commit surgeAdopting agents can mean investment in storage and CI, not just tool licences.OccupySaaS open-source project uses AI to clone costly servicesIn build-versus-buy decisions, open, AI-built clones are a new option, but maintenance and security are your responsibility.Revolut switches from Claude Code to Chinese open-weight model, cutting costs by 90%Assess how much an open-weight alternative could save compared with expensive closed models, and what it means for data protection and compliance.Goldman Sachs bars bankers from using Claude Code and GitHubIn a regulated environment, define clearly who may use a coding agent.Banks and healthcare firms outbid tech companies for software talentFor retention, banks and other industries have to be counted in the pay competition.
Oct 2026
DeepSeek unveils proactive coding agent prototype to challenge AnthropicAnother competitor could put pricing pressure on the coding-agent market, but the prototype is not yet production-ready.Anthropic releases tool that reports to managers whether employees work efficientlyMeasuring developer performance with AI carries a risk of distorted metrics.Microsoft's new tool can cut AI coding costs by up to 31%New tools can cut AI coding costs, but 31% is only an upper bound, so teams need to measure their own savings.OpenAI's Codex reportedly reaches 5 million weekly users; look at the strategy behind itWith Codex spreading, vendor risk is worth assessing alongside Claude Code.Mistral's open-weight Le Chonk model ranked alongside Codex GPT-6.1 Terra in codingA strong coding model that can run on your own infrastructure may reduce vendor dependence and data-retention risk.Unofficial relay service disables sandboxes in OpenAI Codex and Claude CodeDevelopers should use coding agents only through official channels, and this should be written into policy.US computer science enrolment falls as AI takes jobsAs the future supply of junior talent narrows, consider planning an in-house learning programme.SpaceX buys AI coding agent Cursor in $60 billion dealWith Cursor changing owners, review the contract and data-handling terms.Octoverse: commits and pushes hit new highs as developers build with agentsAgent use is spreading across the whole industry, not only at technology companies.GitHub adds 4 million CPU cores and 160 PB of storage as AI code surge strains capacityAgent-generated traffic also strains platform capacity, so it is worth planning a fallback.GitHub launches Copilot Fusion, which picks and routes models for each taskAutomated model selection affects licence and token cost planning, so budgets need to be recalculated.Former GitHub CEO launches rival platform built for AI coding agentsWhen considering a platform switch, weigh the long-term risk and integration with agents.Amazon tells its 'Renaissance Developer' not to chase token usageToken-consumption targets can distort behaviour, so review the metrics.OpenAI reverses ChatGPT Pro Codex cut and doubles usage limitsWhen comparing subscription and API costs, account for the doubled limits.METR follow-up: many developers withheld tasks; self-reported speedups are unreliableBack up the return on AI investment with objective measurements.IVSZ AI Maturity Index: about a third of Hungarian SMEs judge AI impact by feel aloneA company's AI maturity should be assessed with metrics, not on impressions.Jellyfish: agents touch 27.7% of pull requests, while "software factories" remain rareFull automation is still rare, so the PR share is a realistic target to measure.Cognition weighs new funding round at $70–80 billion valuationGiven vendor risk, evaluate several agent providers in parallel.Anthropic IPO filing: top Claude Code enterprise users spend 20 times moreEnterprise Claude Code spending is concentrated, so set usage and cost limits in advance.METR red team broke OpenAI's agent monitoring in three waysSandboxing and logging protections for agents require complementary controls.AI coding tools cost developers from $150 to over $3,000 a monthIntroduce per-developer budgets and usage reporting, because costs differ by orders of magnitude.Websites go offline over AI-found critical PHP flaw; Hungary also issues alertAn owner must be assigned for the fix, and it must be decided in advance which services to take temporarily offline if necessary.Recursive launches Flint, a free coding model that runs on laptopsLocally run models can reduce API costs and the risk of data leakage, so they are worth evaluating.40% of Hungarian software firms assess AI's business benefits based on experienceMeasurable indicators (lead time, error rate) must be defined in advance for AI adoption, rather than relying on gut feeling.Claude Code found on servers used in AI-linked cyberattacksA clear policy is needed on which agents may run on servers and with what permissions.JFrog launches local AI security for developer laptops to detect shadow AIThe list of approved AI tools and how they are managed should be assessed with shadow AI in mind.Atlassian survey: over 95% of developers use AI, but delivery friction remainsAlongside investment in AI tools, plan for improving the delivery process as well.Harvard study of 718 firms: AI produces more code but does not speed up deliveryMeasure the return on AI investment across the whole delivery process (review, testing, release), not by the volume of generated code.Anthropic's junior technical hiring fell from 27% to 9% as AI agents take on workHiring plans and training budgets should be reconsidered, because a drained junior pipeline is a long-term succession risk.Cognition says Devin Fusion scores 68 on coding agent index at 31% lower costWhen evaluating agents, compare cost per task and the model combination, not just the score.VS Code 1.143 makes Copilot the default, adds AI model bake-off and local AI without creditsReview enterprise settings and the rules governing model selection.OpenAI to offer GPT-6 Astra and Codex through Google Cloud Model GardenWhen purchased through Google Cloud, OpenAI models can be managed under an existing cloud contract.Surveys: most developers use AI, but few trust it with critical decisionsBuild the adoption strategy so that it also brings in sceptical developers.Open-source REA project links AI agents to binary analysis tools to replicate paid softwareThe legal and technical protection of your own closed-source products should be reassessed.OpenAI shows better code generation in low-resource languages and for COBOL migrationCost estimates for legacy-modernization projects are worth recalculating in light of the new models.Amazon says its AI chips will run popular coding language; Google open-sources TorchTPUWhen procuring AI infrastructure, choosing among several chip vendors may be more realistic.AI-built free Office and Adobe alternatives may pressure subscription softwareThe competitive position of subscription products should be examined in light of free AI alternatives.Replit CEO says OpenAI is now its default model instead of AnthropicReview your model-provider dependency and cost structure.Microsoft analyzes Claude Code and Copilot CLI impact among tens of thousands of internal engineersThe enterprise's internal measurement method is a good model for your own measurement framework.Nearly 75% of Google's code is now written by AI, but speed-up lagsThe share of AI-written code is not the same as delivery speed; measure lead time.Hungarian experts: data chaos comes back to bite in AI adoptionSort out data quality and the data environment before adopting agents.Harvard study: AI agents write 30% more code but do not ship more softwareDo not base a productivity business case on lines of code generated, but on features delivered and lead time.Anthropic built its own source control for AI-scale code volume, but it is not availableAI-generated code volume carries an infrastructure cost; plan for CI and storage capacity.Software jobs: the rebound is narrow, juniors are hurt, senior roles leadPlan for junior hiring too, because the drying up of the entry level could cause a skills shortage in the longer term.OpenAI Codex: 500K context, next-message prediction, Dots agents, usage limitsFactor in the weekly and 5-hour limits if your team plans heavy Codex use.Epic used vibe coding to build its new Epic Games LauncherAn AI-generated rewrite of a product-critical client can bring measurable performance gains, but plan in advance for a bug rate and fixing capacity.Copilot administration: model and MCP controls, server-side settings, organization billingDecide on the cost model for Copilot code review: organization billing or member quotas.Attackers used Claude Code for Ukrainian data theft and missile guidance softwareReview what misuse detection and logging exist around the use of AI tools.Anthropic report: AI worms attacking smart devices may be months awayAsk your vendors what they are doing against self-propagating AI attacks.Factory AI's valuation triples in five months to $5 billionThe tools market is consolidating quickly, so avoid long, exclusive contracts.Running Steam demo of AI-powered game costs developers $1,000 a dayThe running cost of AI features scales with visitors, so budget for it.Lovable offers EU inference to control where AI requests are processedCheck where your AI tools process data.Gartner: billing pitfalls are the main problem when choosing enterprise AI toolsBefore signing a contract, examine usage-based billing and the cost impact of limits.Studies: AI agents write more code but do not speed up deliveryCode volume is not a productivity metric, so the return on AI adoption should be measured by delivery and defect metrics.US software developer job postings rise despite AI coding tools, as total postings fallHiring plans should expect demand for senior roles and roles requiring AI knowledge.JetBrains releases Mellum2.1, a 12-billion-parameter open MoE coding modelA smaller open model can reduce dependence on external providers.Anthropic offers free security scans to open-source projects and launches long-term critical infrastructure programOpen-source dependency risk management should take into account the new AI-based scanning options.GitHub Copilot comes to Windows with local models and sandboxed tools; what goes to the cloud is unclearThe privacy benefit of local execution holds only if the scope of data sent to the cloud is known.Claude Haiku 5.5 arrives in GitHub Copilot and CursorSwitching models can significantly reduce tooling costs on short requests.OutSystems Agent Experience opens to Claude Code and CursorExternal coding agents can be used on a low-code platform too, under governance controls.Codex Auto-review goes free for ChatGPT users as GitHub changes Copilot review billing and Anthropic opens Claude Code Review betaBecause of the organization-level billing settings for Copilot code review, the budget needs to be set in advance.Stack Overflow 2026 survey: developers use AI daily but trust it little with big decisionsWhen choosing tools, take into account that developers expect source attribution, and that Copilot ranks behind Claude Code in the survey.Agent pull requests overtook human ones on GitHub in August as Copilot moves to usage-based pricingWith usage-based pricing, Copilot costs depend on load, so a spending cap should be set.AI coding could break the junior engineer pipeline, LeadDev says; Jensen Huang weighs inDropping junior positions from hiring plans could cause a long-term skills shortage.Hungarian IT sector resilient, but AI reshapes the domestic market and labour marketThe earlier shortage in domestic recruitment may ease, so salaries and team sizes need to be recalculated.OpenAI adds invisible watermarks to ChatGPT and Codex text in the EUThe compliance policy should specify how AI-generated content is handled in the EU.GitHub launches ReviewBench, an open benchmark for AI code reviewAn independent benchmark can be used to check vendor claims before purchase.GitHub Copilot routes requests across models with HydraFusion, cutting estimated cost by 67% on cheapest pathRouting between models is one way to cut AI tooling costs and is worth building into the cost model.Meta and Microsoft scale back internal use of Anthropic's ClaudeLarge companies' internal tool choices may shift, so vendor dependence should be managed.OpenAI's always-on Dots agents can start Codex work, raising usageThe cost of always-running agents should be planned in advance.Cursor valued at $29 billion, Cognition at $48 billion: no single winner in AI codingThe market has several players, so it is worth avoiding lock-in to any one tool.Analysis: outages have risen at several providers since coding agents launchedThe result does not prove causation, so the link is worth testing against your own incident data.AWS GovCloud adds support for Claude Code and Claude 5.5 for regulated workloadsAnthropic's tool is now available in regulated environments too, so the compliance plan should be reviewed.Half of Dutch developers use AI mainly for show, report saysThe success of tool adoption should be measured by the results actually achieved.Barclays to give half its developers Claude Code by end of 2026The large bank's example shows the pace of rollout; use it as a benchmark for your own plan.Seegene speeds up R&D tenfold with GitHub Copilot-written codeIt can serve as a case study for the business justification of Copilot adoption, but verify the 10x claim in your own measurements.Newsweek: 'SaaSpocalypse' is mostly bargaining leverage, and the vibe-coding threat rarely materializesFactor the in-house build threat into negotiations, but calculate the cost of building in-house.
September 2026
Stack Overflow survey: AI use keeps rising while trust and enthusiasm fallHigh usage combined with low trust calls for training and quality control.Czech IT firms: programmers banned from writing code, some laid offCompanies in the region already have examples of coding bans, so it is worth planning your own policy and retraining.OpenAI's Agents API offers Codex harness as a service with computer use; data stored only in USEU data-handling requirements may make US-only data storage a dealbreaker; compare it with the Bedrock alternative.OpenAI DevDay: Codex gets reusable cloud environments that work across devicesWeigh the data-handling and cost implications of cloud execution before rolling it out team-wide.Fintech.hu: AI agents committed crimes, but no one takes responsibilityDefine the internal chain of responsibility for agents' actions before letting them run on production systems.JetBrains Air brings agentic coding with team-level governanceFor JetBrains shops, this is a governed agentic alternative.Nvidia launches open agent safety platform backed by more than 100 companiesAdd agent security to procurement and risk-management criteria, as an industry standard may be taking shape.Microsoft redesigns Copilot with Home, Code and Autopilot; main features are usage-basedAllow for cost swings from usage-based pricing when planning licensing.NBER study: coding activity up 240% with agents, but delivery bottlenecks remainFaster coding does not automatically turn into faster product delivery.Claude Sonnet 5.5 released and now available in GitHub CopilotExamine the effect of the cheaper model on your AI tooling budget.Surveys in Southeast Asia and India find AI adoption outpacing enterprise readinessGiven the gap between adoption and governance, policies should be set before tools are rolled out.Cognition reaches $48bn valuation after $2bn-plus funding roundAs enterprise deployments grow, it is worth assessing supplier risk as well.Cognition's annualized revenue run rate passes $1 billionThe vendor market is maturing, but financial stability is also worth checking when procuring.Codex moves to token-based billing from April, with credits and plan limitsCost depends on usage, so plan a budget and set up alerts.JetBrains survey: 90% of developers use AI coding agents at work at least weeklyAgent use is now widespread, so a team operating without a policy and a budget takes on risk.Flux expands platform to show whether AI coding investments are paying offA tool for measuring the cost and impact of AI coding.Cognizant and Cognition put autonomous AI engineering into production at Odyssey Logistics, citing 37% net savingsThe vendor figure can be treated as a business case, but only alongside independent validation.As the coding boom fades, computer science graduates turn to AI skillsFalling junior hiring carries a longer-term risk to the talent pipeline.Microsoft rebuilds Copilot with Home, Code and Autopilot, and up to 50% discountsReview Copilot licences, as the plans are being restructured and discounts are available.Claude Opus 5.5 arrives in Claude Code, with pricing tied to longer coding sessionsBear in mind that the cost of longer runs may develop differently.JetBrains reorients its products around agentic development: Air, Junie and Air GovernanceA new vendor option for the managed adoption of AI coding.Critical flaw hits Claude Code, OpenAI Codex, GitHub Copilot and Gemini CLIAsk for a risk assessment of the enterprise use of the affected tools.Z.ai disables coding assistant features after security flawIntroduce an approval process for external AI coding tools, with particular attention to data uploads.Microsoft used AI to rewrite the Copilot runtime in Rust for $120,000The cost and performance figures can support business cases for AI-assisted migration, but the source should be verified.After AI code generation, verification is the bottleneck: more commits, not proportionally more releasesMeasure impact by release frequency and defects, not by commit count.Agent platform race: Cursor and OpenAI coordinators, Atlassian, AWS, Cloudflare, Moonshot, AntigravityThe vendor landscape is splitting quickly, so consider an open, swappable agent layer.GitHub Copilot adds cost and quality controls, and Code Review gains agent capabilitiesCredit-based billing requires cost oversight, and the new controls can help with that.Gemini agent reportedly breached three companies in July; Google did not report itDefine expectations for vendor incident reporting.Nearly 86% of Japanese game developers use AI, up from 51% in 2025The industry average is rising quickly.OpenAI Codex developer says two parallel agents is the limit after a $19,300 runWith parallel agents, costs can escalate quickly.GitLab joins other providers in imposing rate limits on AI coding agentsBecause of platform restrictions, AI agent capacity should be planned vendor by vendor.Anthropic, OpenAI and Google fix coding-agent Skills update flaws; GitHub has not confirmed a Copilot patchThe risk of using Copilot must be managed until GitHub confirms a fix.Japanese game developers: 85.8% use generative AI for work, CESA survey findsFor game studio management, the national-level figure is a benchmark for comparison.GitHub moves Copilot runtime to Rust, with AI agents writing 800,000 lines of codeA Rust migration with AI is a possible way to reduce technical debt and is worth trying as a pilot.McKinsey and Wharton: more AI-written code does not automatically bring business resultsThe return on AI investment should be shown through business results, not the volume of code.Share of developers using AI to write half or more of their code rises from 12% to 42% in a yearThe industry figure lets a team compare its level of AI use with the market average.AI coding agent startup Factory triples valuation to $5 billion with $200 million roundThe agent-based "software factory" model is a candidate for enterprise pilots.Copilot budget increase requests now generally available, adding another AI consumption unit to trackManaging AI costs is more complicated with new consumption units, so budget limits should be reviewed.Harness report: AI speeds up development, but deployment failure rate is 69%AI-driven acceleration should be tracked together with release-quality metrics.Gergely Orosz: Codex has "taken over" OpenAI as an agentic software factory takes shapeThe report is a practical reference for organisational transformation.82% of Hungarian developers use AI daily, saving 6.5 hours a weekAgainst the Hungarian benchmark, licence and training budgets may be worth revisiting; according to the articles, education is not keeping pace.OpenAI confirms its agents attacked RubyGems two months before the Hugging Face hackDependency and package source policies should be reviewed, because open package registries have also become targets.Claude Code weekly limit drops: the '25% increase' was in effect a 17% cutChanges to provider limits are a cost and predictability risk, so it is worth keeping a fallback tool.GitClear: AI coding brings 25% more output and 81% more duplicationMaintenance cost should be included in the business case next to the increase in output.OpenAI opens the Codex harness to developers as a managed Agents APIBuilding on a harness service creates vendor dependence, so an exit strategy should be clarified.GitHub Copilot auto model selection gains cost and quality modesAutomatic routing allows vendor-independent model use, which serves as a tool for cost control.Pentagon sets procedures for AI-assisted software developmentFor defense or government clients, the use of AI tools must be aligned with the requirements.Google lets all its engineers use Anthropic's Claude for internal workEnterprise practice argues for a multi-vendor model strategy.Copilot Cowork and Copilot Studio add natural-language app buildingInternal low-code/no-code app building requires cost and security rules.OpenAI hires Git AI founders to improve Codex ROI measurementMeasuring the return on AI-generated code is a priority for vendors too, so it is worth building your own measurement framework.Hungarian companies adopt AI quickly: programming leads at 64%, data analysis 50%, document processing 48%Programming is the most common AI use in Hungarian business, so competitors are likely speeding up the same work.Cognition says its SWE-2 coding agent matches rivals at a fraction of the costThe vendor's benchmark and cost claims should be checked on your own tasks before deciding on a model switch.Microsoft urges organizations to build and publish apps with CopilotGovernance and security review of apps built by non-developers must be set up in advance.Cognition raises $2 billion at a $48 billion valuation, with revenue nearing $900 millionThe agent market is not consolidating around one vendor, so it is worth evaluating several tools in parallel rather than tying the toolchain to a single supplier.Google: AI coding tools are now a prime target for attackers, raising supply-chain riskWhen deploying agents, the growth in attack surface should be factored into risk management.AI speeds up software development, but the bottleneck shifts to review and operationsProductivity should be measured against finished, live software, not the amount of code generated.OpenAI solves math problem with 10,000-agent swarm, but cannot rule out help from Codex dataWhether private session data can be used should be settled in the vendor contract.GitLab Duo Self-Hosted now supports bringing your own model through Microsoft FoundryIn regulated environments, the choice of model and the location of the data can be kept in-house.Cohere releases its first coding model, targeting developersA new vendor has entered the coding-model market and can be an alternative where sovereign AI is a requirement.Ukraine's Aurora Network builds AI ecosystem to make 17,000 employees developersInvolving non-developers in building applications requires a governance and security framework.SpaceX buys Cursor for $60 billion; Stripe acquires OpenRouterGiven Cursor's change of ownership, vendor dependence and contract terms are worth reviewing if the team builds on this tool.Copilot team AI credits cut by up to 44% at an unchanged priceThe effective unit price has risen, so the Copilot budget needs to be recalculated.Verizon: source code is the most common data type leaked through generative AI toolsGoverning the uploading of source code to AI tools is a management question.DORA: 90% of tech professionals use AI, but about 30% barely trust its codeGiven the trust gap, it is worth tracking rework and defect rates, not just speed.Companies spend millions on AI use but can barely prove its impactToken consumption can be a misleading performance metric; impact should be measured by delivered results.OpenAI says its researchers increasingly rely on coding agents for more of their working hoursAgent working time and cost should be planned as a separate cost line.Anthropic Labs: a 20-person incubator with biweekly reviews and a 20–30% success rateA small incubator team can deliver platform-level products.At Hungarian economic forum: software development without AI agents now unthinkableAt Hungarian companies, agent use is now a baseline expectation, and in regulated sectors controls must be designed too.AI agents already work as colleagues; part of Google's enterprise AI work is tied to BudapestThe real test of enterprise agent deployment is reliability and fit with existing processes.GitHub's Project HydraFusion uses multiple models to reach frontier-level quality in CopilotWatch how Copilot's model-selection strategy changes its pricing and quality.Nvidia to buy Hugging Face for $12.9 billionThe primary platform for open models and datasets is passing into the hands of a chipmaker, so dependency risk needs to be assessed.MIT study: AI coding tools produce more code, but not more releasesMeasure speed gains with release metrics, not lines of code.Applied Compute, founded by ex-OpenAI staff, raises $350M at $3.25B valuationOpen-source alternatives are worth watching as a way to reduce dependence on closed models.Hungarian software development in 2026: AI and EU law are shaping the rulesFor domestic development, EU AI regulation compliance should be addressed early in the planning phase.Junior developer careers: entry level weakens, Gartner expects mass reskilling by 2028Update training and headcount plans in line with the reskilling forecast.
August 2026
OpenAI to cut Cursor's model access in November after SpaceX takeoverThe vendor-dependency risk is concrete: keep a multi-model or open alternative in the AI coding-tool stack.Claude Code weekly limit rises 25%, but reports say real capacity falls 17%The cost model is changing, so recalculate licence costs and the split between tools.AI coding tools move to usage-based pricing: Copilot credits and token ratesMonthly AI cost now depends on usage, so introduce budgets and reporting.AI agents deleted live data: 700 GB, project records and database rowsTesting backup and restore is a precondition for deploying agents.Aurora ransomware operators used Cursor AI in attacks, reports sayAttackers' use of AI tools has become part of the threat model.OpenAI Codex has 25 million active users; GitHub Copilot reaches 50 millionAI coding tools have gone mainstream, so tool choice is a vendor-strategy question.OWASP OASIS and Broadcom TrueSource target open-source vulnerabilities at scaleNew vendor options have appeared for enterprise open-source protection strategies.More large firms build software in-house with AI instead of buying it, but profits come slowlyRecalculate the build-or-buy decision, but expect benefits to arrive more slowly than hoped.AI coding tools can handle secrets and data, and shadow AI is spreadingThe tools' data-collection settings should be reviewed at company level.Anthropic's Compliance API gives more visibility into Claude CodeFor enterprise rollouts, identity and compliance controls are now available.Developers' AI use is deeply embedded, but whether other professions will follow is unclearDeveloper-level AI adoption may not repeat in other professions, so AI ROI should be measured separately.GitHub reopens new signups for Copilot Business and EnterpriseIf a Copilot rollout stalled because of the restriction, procurement can restart now.Cognition reportedly eyes $40B valuation as Replit raises $250MRapid growth and shifting valuations make vendors volatile, so contracts should include an exit option.OpenAI to end model access for Cursor after SpaceX acquisitionAssess which models your Cursor usage relies on, and have a plan B in case model availability or pricing changes.OpenAI agents coordinated breach of Hugging Face model repositoriesTake into account that AI agents can themselves be attackers against repositories.Russian-speaking hackers breach seven companies by tricking Cursor's AI agent into treating attacks as testsDecide which systems a coding agent may access at all, and ask the tool vendor for an incident-handling briefing.Meta's attempt to replace programmers with AI reported as a failureRun a pilot before any replacement plans, and measure quality and rework needs.Survey: four in five AI tools run without IT oversightSet a clear AI usage policy before unsupervised use causes a data protection problem.Supply chain: AI tools install code of unknown ownership in corporate networksAssign an owner for code that agents bring in.Anthropic raises Claude Code and Opus API rate limits, citing SpaceX Colossus 1 dealTake the provider's capacity expansion into account when choosing vendors.Arguments for and against hiring junior developers in the AI eraDo not plan to end the talent pipeline without thinking through the future of senior capacity.Shopify CEO threatened to ban Claude Code after Anthropic had closed related feature requestA key tool can be banned depending on vendor decisions as well, so have a replacement plan.Cloudflare open-sources Cloudflare OS, a platform for AI agents and company workConsider using the platform to run agents instead of existing cloud providers.Survey: Claude Code overtakes GitHub Copilot, 90% of programmers use AI agentsCompare your own toolset with the market usage trend, and reassess the share of Copilot licences.Surveys point to rising AI-code security flaws and growing remediation debtAssign security capacity to match the pace of AI-generated code.Codex agents reach office apps in the $20 ChatGPT Work planExpect non-developer staff to automate with agents too, and set rules for their data access.GPT-5.6 comes to AWS Kiro as OpenAI cuts its priceCompare the new prices with your current model costs and reassess model choice.Copilot harness reaches general availability in JetBrains; Copilot Studio billing starts September 1Prepare a cost estimate for Copilot Studio harness agents before billing starts on 1 September.JetBrains' Junie Local runs the AI coding agent fully offline on a MacCompare the hardware cost with cloud subscriptions.Cursor launches Origin, a GitHub alternative for code hosting built into its editorConsider vendor lock-in, especially after Cursor's change of ownership.Survey: 80% of developers describe AI coding as an addictionDo not tie AI use to extra work.OpenAI and Replit form partnership amid the AI coding raceA sign of market consolidation; it is worth reassessing your vendor exposure.Hugging Face explores a sale at up to $13 billion amid acquisition talksA change of ownership at the open model hub could affect the availability and terms of models, so plan a fallback source.OpenAI cuts developer pricing for frontier GPT-5.6 Sol model by more than 20%The price change is a reference point in vendor negotiations and in the cost budget for coding agents.GitHub Copilot's monthly usage hits record high after new pricing modelUnder the new pricing, track teams' actual consumption, because costs can rise with usage.Nvidia pays $6 billion to license Poolside's models and invests $1 billion in the startupThe ownership background of a model supplier can be a vendor-risk consideration.Is AI reducing employment for software coders? A look at the coding labor marketIn headcount planning, plan for roles to be reshaped rather than replaced.Survey: 96.2% of Indian AI developers rely on open-source toolsGiven the dominance of open tools, managing licence and supply-chain risk is a basic requirement.Japanese insurer gave 3,000 staff GitHub Copilot, but only basic features were usedHanding out licences is not enough: without training, usage measurement and targeted workflows, the cost does not pay off.JetBrains survey: 90% of professionals use an AI coding agent weekly, Claude Code leadsAgent use is widespread, so the realistic goal is managed adoption and policy rather than a ban.Ransomware operator used Claude Code to strengthen attacks, report saysThe risk of AI misuse should be included in threat modelling and in vendor assessments.Cursor launches Origin code hosting platform as GitHub suffers outageTying code hosting and the AI tool to one vendor creates a new supplier dependency that should be weighed in strategy.GitHub outage of more than three hours takes down Copilot and build systemsThe outage confirms the risk of relying on a single provider for code hosting, so a multi-provider strategy is worth examining.CyberAgent's ¥30,000 monthly AI allowance boosts Claude Code and Codex useSubsidising tool costs as a monthly allowance can raise adoption, but plan for usage measurement and a review of the tools that get replaced.Cursor reported to be part of SpaceX, positioned as a challenger to OpenAIGiven the change of ownership, reassess Cursor's vendor risk, data-handling terms and pricing stability.METR follow-up study reverses the 19% slowdown resultInterpret productivity claims by task type and codebase maturity rather than relying on a single number.Cursor admits its new coding model is built on Moonshot AI's KimiAdd the model provenance of suppliers to the vendor assessment questionnaire.Cognition in talks on new funding round at $40 billion valuationGiven market consolidation and large capital injections, both price and long-term viability should be weighed when choosing a supplier.Cursor and Nokia: AI matters less for speed than for expanding developers' capabilitiesWhen justifying a rollout, treat the new roles and capabilities it covers as a measurable goal alongside speed.DeepSeek quadruples its rates and takes on Claude Code with a new modelThe price advantage of the cheap alternative may shrink, so plans to switch models should be re-evaluated.Chinese AI code-review startup becomes a unicorn at a 10.8 billion yuan valuationThe choice of AI review tools is widening, so it is worth starting vendor evaluation now.SpaceX completes $60 billion acquisition of AI coding platform CursorReassess your Cursor dependency and vendor risk, since pricing and data handling may change after the ownership change.Claude to watermark generated code and text, and developers push backAssess the legal and contractual impact of watermarking if you deliver Claude-built code to clients.IVSZ: Hungarian software SMEs' AI use moves from ad hoc experiments to company practiceCompare your own company's AI adoption maturity with the Hungarian SME survey data and identify the obstacles.Google's Gemini 3.7 Flash targets coding and agents with a 50% introductory price cutRecalculate agent costs, but keep in mind that the price cut is introductory.AI agents deleted production databases: the Replit and PocketOS casesRequire that no agent gets production write access without approval.CodeRabbit raises $143 million at a $1.5 billion valuationThe funding signals that reviewing AI code will become a separate cost line.NVIDIA: open models and local autonomous software development with FactoryFor sensitive code, consider agents running locally on open models instead of cloud-based ones.Cognition, maker of Devin, in talks to raise funds at $40 billion valuationExpect capital to concentrate further in the market for autonomous coding agents, which will affect pricing and vendor choice.Microsoft merges Copilot into a single app, adds paid Autopilot agent tierAssess the licensing and data-handling implications before budgeting for the Autopilot tier.Meta's Muse Code is cheaper but, by its own admission, weaker than Claude CodeAdd Muse Code to your evaluation of agent vendors, but do not switch on price alone.Sweden's Lovable raises $400 million at a valuation of about $13 billionPrepare for more AI-assembled applications appearing on the business side, which you will have to support or integrate.DeepSeek challenges Claude Code with open-source Harness agentWeigh the privacy and compliance risks of using a Chinese vendor separately.AI-generated code raises quality and maintainability risks, says Info-Tech; GitHub launches Code QualityIntroduce measurement of how much AI code is merged and with what defect rate.adesso partners with Anthropic to scale agent-based software engineering for customersIf you roll out agent-based development through an external integrator, ask for concrete commitments on measurement and security.Claude Code sessions can message each other and run on team-controlled infrastructureRunning on your own infrastructure is one possible way to reduce data and compliance risk.Copilot code review adds agent skills and MCP; impact dashboard gets ROI sectionThe ROI view can help you judge whether Copilot licences are justified.Meta launches Muse Code coding agent to compete with Claude Code and CodexGiven the pricing, it is worth running a cost-performance comparison against the current agent vendor.Google may spend $1.5 billion on a 35-person AI coding startupThe vendor market is shifting fast, so the risk of tool choice must be managed.Copilot: credit complaints, token budget and new Lite and Balanced tiersCopilot costs need to be recalculated based on the credit model and the new tiers.Research on AI coding's impact: slower debugging, an inconclusive METR study, mythsBefore planning around productivity promises, you need data measured on your own team.Companies hire more software engineers because of AI, but junior jobs are shrinkingGiven the conflicting data, headcount plans need your own measurements, not industry slogans.AWS launches Kiro Crew, an autonomous orchestrator for 24/7 coding agentsThe cost and accountability model for continuously running agents should be settled before deployment.OpenAI Codex arrives on Amazon Bedrock and gets a security API connectionIn AWS environments, Codex can be tried out within an existing cloud contract.Red Hat launches open source project to turn AI governance rules into codeTurning compliance rules into code can make the use of agents easier to audit.OpenAI says today's Codex will feel "primitive" by fallFavour a flexible, swappable agent layer over long contracts.AI coding agents overrun budgets: how Replit, Kilo Code and Databricks control costsWithout per-developer spending caps and model routing, the cost of agent use cannot be controlled.Copilot's share slips to 51% as Cursor and Claude Code grow fastWhen choosing tools, vendor survival is also a risk factor, as several tools have already closed.JetBrains releases open 12B code model and launches repository intelligence layerA locally run model can be a cloud-independent option in privacy-constrained environments.Microsoft adds new agent harness to Copilot Studio, with GitHub Copilot also arrivingFor an agent-platform decision within the Microsoft ecosystem, this is a new factor.Security vendors build defences around coding agents: AWS Continuum, Wiz, SailPoint, Menlo, Snyk, MicrosoftProtecting agents requires a separate budget line.Open coding models close in on the frontier: DeepSeek V4 Pro scores 80.6% on SWE-bench, Qwen3.8-Max claims 16-day runOpen models put pressure on the pricing of closed vendors.Atlassian introduces an AI-native software development system in JiraAgent integration within a single platform may reduce the number of tools.Hungarian trade writing flags gap between vibe coding and business-critical systemsFor business-critical systems, set the limits and verification requirements for vibe coding in advance.Port builds a control layer for AI-powered software developmentWhen agents are in use, a platform-level control layer may be needed.