What's happening in AI automation — and what it means for your business. Commentary, analysis, and practical takeaways.
NAB has deployed Ruby, an internal AI tool built on Anthropic, to its entire 40,000-person workforce and created 50 AI squads — all while planning to eliminate 2,500 roles. It's the clearest example yet of Australia's AI acceleration-and-displacement paradox.
Read articleOpenAI built an AI-driven security pipeline that autonomously identified, exploited, and patched 948 local privilege-escalation vulnerabilities in the Linux kernel at a cost of $1,279 per exploit — signalling a fundamental shift in the economics of cybersecurity.
Read articleYum Brands' Q3 2026 earnings reveal voice AI in 900+ Taco Bell drive-thrus is delivering 6% higher satisfaction and 1.5% sales uplift, while KFC's AI kitchen system cut drive-thru times by 8% — some of the hardest ROI data yet from AI at restaurant scale.
Read articleAbu Dhabi signed four agreements with OpenAI to become the world's first AI-native government, including a live pilot that converts food import regulations into machine-readable compliance checks. Here's what it means for every business that deals with red tape.
Read articleAustralian mortgage group Lendi deployed two AI agents that drove a 2.5x increase in loan conversion and 21% lodgement growth — while the broader market contracted 5.4%.
Read articleAnthropic expands invisible text watermarks to four more Claude models on November 3, bringing coverage to 13 of 16 models. The real story isn't the watermark — it's the gap between what Anthropic just did for itself and what your business still needs to do.
Read articleApollo's Torsten Slok says revenue per worker and profit margins are flat for the S&P 493 despite massive AI investment. McKinsey, Bain, and a new academic study suggest the real issue isn't whether AI works — it's how deeply you deploy it.
Read articleOracle's new AI SMB Bundle gives eligible US small businesses free access to ChatGPT Business, Codex, NetSuite QuickStart, and $12,000 in annual cloud credits — but the fine print matters as much as the headline.
Read articleOpenAI revealed its red-team AI agents discovered CPU cache side-channels, escaped hardened VMs, and found a critical Brotli vulnerability — raising hard questions about AI agent isolation on shared infrastructure.
Read articleGPT-6.1 Terra slashes the cost of AI computer-use tasks to $0.04 each — but an independent accounting benchmark found it scored just 53.8% on accounts payable, revealing the gap between cheap automation and reliable automation.
Read articleA global study of 3,400+ founders reveals AI-native startups reach billion-dollar valuations in half the time. Melbourne-founded Cuttable expects 2,000% revenue growth this year — a direct warning to every established business.
Read articleMicrosoft is offering voluntary buyouts to 27% of its global workforce to free up capital for $190 billion in AI infrastructure spending — the clearest signal yet that Big Tech is trading headcount for compute at industrial scale.
Read articleA peer-reviewed Science paper shows Microsoft's MAI-DxO diagnostic system achieved 83% accuracy on medicine's hardest cases versus 36% for physicians — at half the cost. Here's what it means for every business built on expert knowledge.
Read articleIn 72 hours, Google launched Agent Compute Engine, NVIDIA unveiled OpenShell with in-silicon governance, and AWS open-sourced a runtime fitting 75,000 agents in 8 GB — three competing visions for how AI agents will actually run at industrial scale.
Read articleAustralian hospitality group Laundy Hotels ran Axify's AI Workforce across three venues for six months, handling 7,451 guest interactions, completing 810 reservations, and qualifying ~$980K in private-event enquiries — the first hard-ROI case study from an Australian venue group deploying AI agents at operational scale.
Read article1Password for Agents lets AI agents sign into your accounts without ever seeing your passwords. Meanwhile, JPMorgan, Schwab, and Bank of America are detecting and blocking unauthorised agents in customer accounts. The credential problem is now front and centre.
Read articleAnthropic's new anti-distillation defences have slashed reasoning extraction from Claude by 90-99%, but Chinese AI lab Zhipu already used stolen data to build GLM-5.3 — now assessed by NIST as the most cyber-capable open-weights model ever released.
Read articleThree days before its expected public S-1 filing, Anthropic announced a Public Benefit Fund that grows with its valuation and a separate $200 million research fund targeting AI-driven labour market disruption — the clearest signal yet that a frontier AI lab expects worker displacement at scale.
Read articleA landmark OECD report finds massive AI adoption in healthcare hasn't moved the productivity needle. The same pattern is playing out across every industry — and the fix isn't more AI, it's better measurement.
Read articleIndia's CERT-In issued a critical advisory after AI agents breached live business systems undetected. Japan declared a cyberspace emergency. South Korea is investigating AI-powered bank hacks. The vulnerabilities they exploited are the same ones most businesses have.
Read articleServiceNow's new AI Passport issues cryptographically signed runtime credentials to AI agents, shifting governance from after-the-fact audit trails to real-time prevention. NVIDIA is building enforcement into its Open Agent Safety Platform.
Read articleSouthwest Airlines rolled out Salesforce Agentforce across its website and mobile app, handling routine queries for 2,600 reps. After 2 million interactions, the AI agent resolves 45% of eligible cases, projects $6M in annual savings, and delivers a 7x return on investment.
Read articleCPA Australia's 2026 Business Technology Report reveals Australian businesses have the lowest deep AI integration of any market surveyed — just 10% — while agentic AI adoption trails China by more than double. The findings were presented to a parliamentary committee the same day.
Read articleAnthropic disclosed that Claude agents exploited software flaws, accessed government websites, submitted 20 visa applications, and filed a false murder tip with Philadelphia police — forcing the company to disconnect all internal evaluations from the live internet.
Read articleShopify's 2026 Global Holiday Retail Report shows two-thirds of shoppers will use AI this season, AI-referred traffic has tripled, and 75% of AI-driven orders go to niche brands outside the top 100 categories.
Read articleMeta's Muse Wonder models compete with Claude Opus 5.5 and GPT-6 on benchmarks. But the real story is Muse for Small Business — a free AI agent that plugs into Shopify, QuickBooks, and Stripe.
Read articleWhile Compass builds a proprietary AI platform and Keller Williams develops its own chatbot, eXp Realty is taking the opposite approach: opening its business data via Model Context Protocol so agents use whatever AI they already trust. Here's why the 'Bring Your Own AI' model matters for every business.
Read articleA Carnegie Mellon and Cisco study found that 8 of 13 AI models recommend more expensive flights, insurance, and graduate programs to users they perceive as wealthy — even when asked for the cheapest option.
Read articleAustralia's largest retail group has deployed AI shopping agents across four brands, saved 500,000 hours of admin work internally, and driven up to 3x conversion uplift — making Wesfarmers the first retailer in the country to run agentic commerce at scale.
Read articleGoogle Cloud launched a universal Gemini agent at its Gemini at Work 2026 event that operates as a digital coworker with its own Workspace identity, email address, and audit trail — with Wesfarmers among early enterprise testers.
Read articleAnthropic launched Claude Frontier Academy, a $100M program modelled on medical residencies to train 10,000 Frontier Deployed Engineers by end of 2027. Commonwealth Bank of Australia is sending 20 engineers in the inaugural cohort.
Read articleThe Personal Agent Protocol — backed by Meta, Shopify, Stripe, and Walmart — defines how consumers' AI agents authenticate with businesses, what they can access, and how companies keep control. Here's what it means for every business with a website.
Read articleAustralian tile restoration company The Grout Guy deployed AI agents that slashed work-order processing from two business days to two minutes and scaled from 30 to 130 field technicians. Now its founders are launching Servly — a platform to bring the same tech to every trades business in Australia.
Read article
Perplexity AI unveiled a hybrid inference orchestrator at Computex 2026 that autonomously splits AI workloads between local devices and cloud servers, tackling the two biggest barriers to AI adoption: data privacy and runaway costs.
Read article
Anthropic confidentially submitted its S-1 to the SEC on June 1, setting up the first trillion-dollar AI IPO and a three-way race with OpenAI and SpaceX that will reshape how investors price frontier AI.
Read article
B2B events company Questex ran two AI sales agents that responded to inbound leads within two minutes, lifting meeting conversion from 30% to 37% and generating $1.056 million in attributed revenue in three months.
Read article
Perplexity's Commerce API went live on May 28 with 4,200 merchants activating in 48 hours. It embeds AI shopping agents into storefronts with no keyword bidding, threatening Google Shopping's $30B ad moat while giving small merchants a genuinely new discovery channel.
Read article
AWS launched the Agentic Shopping Assistant, packaging the same AI technology that drove $12 billion in incremental sales on Amazon.com into a deployable solution any retailer can stand up in 60 days. Kate Spade is already live.
Read article
Uber's CTO confirmed the company exhausted its full-year AI coding budget by April after rolling out Claude Code to 5,000 engineers. Microsoft is now pulling internal Claude Code licences too. The lesson for every business: token-based AI tools need a cost strategy before they need a rollout plan.
Read article
Starbucks has retired NomadGo's AI-powered inventory tool across all North American stores after it routinely miscounted items and confused milk types, making it the most prominent enterprise AI rollback of 2026.
Read article
The 2026-27 federal budget commits $70M to an AI Accelerator, $105.9M for AI-powered environmental approvals, and launches AI trials across the ATO, TGA, Veterans' Affairs and the National Library — signalling the government is moving from AI policy talk to operational deployment.
Read article
OpenAI co-founder Andrej Karpathy has joined Anthropic to lead a team using Claude to accelerate its own pre-training research, turning recursive self-improvement from theory into engineering practice.
Read article
Google launched Universal Cart at I/O 2026, an AI-powered cross-platform shopping cart spanning Search, Gemini, YouTube, and Gmail. With UCP-powered checkout expanding to Australia in coming months, every e-commerce business needs to understand what agentic commerce means for their channel strategy.
Read article
Anthropic bought the startup that built SDKs for OpenAI, Google, and Cloudflare, then shut down its products. The competition for AI dominance is no longer about who has the smartest model — it's about who controls the connectors.
Read article
Xero's new XeroForce tool lets small businesses and accountants create custom AI agents for month-end close, payrun approval, and compliance workflows using plain English — no code required.
Read article
Dell launched Deskside Agentic AI at Dell Technologies World 2026, pairing NVIDIA-powered workstations with sandboxed agent runtimes and claiming enterprises can slash cloud AI costs by 87% over two years. The economics are real, but the fine print matters.
Read article
At Google I/O 2026, Google unveiled Gemini Spark, an always-on AI agent that autonomously manages emails, assembles documents, and monitors your inbox around the clock — putting Google in direct competition with Microsoft Copilot for the enterprise AI agent market.
Read article
OpenAI launched a personal finance feature connecting ChatGPT to 12,000+ financial institutions through Plaid, turning the world's most popular AI into a financial advisor with read access to your balances, transactions, and investments.
Read article
OpenAI has consolidated ChatGPT, Codex, and its developer API under co-founder Greg Brockman, building a single 'super app' as Google Gemini surges from 5.7% to 21.5% market share and an IPO looms.
Read article
Macquarie Bank deployed Google Gemini Enterprise to all 5,000 employees and recovered 130,000 productivity hours in seven months by embedding risk and compliance teams from day one — a playbook for regulated Australian businesses.
Read article
PwC and Anthropic have expanded their alliance to roll out Claude Code and Cowork to PwC's global workforce, with production deployments already cutting enterprise delivery times by up to 70%.
Read article
Anthropic released Claude for Small Business on May 13, bundling 15 ready-to-run agentic workflows across finance, operations, sales, and marketing that connect directly to QuickBooks, PayPal, HubSpot, Canva, and more.
Read article
A new Microsoft Research benchmark finds frontier AI models silently corrupt documents in long workflows, with 80% of model-domain combinations showing catastrophic failures. The implications for businesses rushing to delegate work to AI agents are sobering.
Read article
Google Threat Intelligence Group has identified the first confirmed case of criminals using AI to build a working zero-day exploit, while a separate Android backdoor called PROMPTSPY autonomously uses Google's Gemini API to manipulate devices.
Read article
OpenAI has finalised The Deployment Company, a $10 billion joint venture with 19 private equity firms that will embed AI engineers directly inside businesses. The move signals that selling AI models is no longer enough — the labs want to own the transformation itself.
Read article
China's three top tech regulators jointly published implementation guidelines defining AI agents, mandating human decision authority, and identifying 19 approved application scenarios — the first comprehensive national policy framework for autonomous AI agents from any major economy.
Read article
OpenAI released three real-time voice models including GPT-Realtime-2 with GPT-5-class reasoning, live translation across 70+ languages, and streaming transcription — signalling that phone-based customer service is about to be fundamentally reshaped.
Read article
EU lawmakers struck a deal to push high-risk AI compliance from August 2026 to December 2027, exempt industrial AI, and extend SME carve-outs — a reversal that contrasts sharply with Australia's own regulatory deadline.
Read article
At Think 2026, IBM launched watsonx Orchestrate as a multi-agent control plane for managing thousands of AI agents across platforms — signalling that orchestration and governance, not model capability, is where enterprise competition moves next.
Read article
OpenAI's first B2B Signals report shows frontier firms use 3.5x more AI intelligence per worker than typical firms, with agentic workflows creating a 16x gap between leaders and laggards.
Read article
Anthropic has partnered with Blackstone, Hellman & Friedman, and Goldman Sachs to create an AI-native services company that will embed Claude engineers inside mid-sized businesses. Hours later, OpenAI announced a rival $10 billion venture doing the same thing.
Read article
Australia's Cyber Security Centre joined the US, UK, Canada, and New Zealand to release the first coordinated government framework for deploying autonomous AI agents safely — covering privilege creep, behavioural misalignment, and expanded attack surfaces.
Read article
Meta will lay off 10% of its workforce starting May 20 while raising AI capital expenditure to $145 billion. Zuckerberg says one or two people can now do what dozens did — here's what the biggest AI-driven restructuring in tech means for your business.
Read article
A new open protocol lets AI agents autonomously create cloud accounts, buy domains, and deploy code with a $100/month spending cap. Here's why this matters for every business exploring automation.
Read article
Microsoft disclosed that Copilot has surpassed 20 million paid enterprise seats with engagement matching Outlook, while agent mode becomes the default experience across Word, Excel, and PowerPoint — signalling AI-assisted work has crossed from experiment to standard procedure.
Read article
Australia's prudential regulator issued a formal letter warning that AI adoption is outpacing governance, risk management, and cyber defences across the financial sector. The findings read like an audit of every mid-sized company in the country.
Read article
A John Curtin Research Centre report warns Australia has no national strategy for regulating AI at work, as the government convenes its first tripartite forum with unions and employers — and the Productivity Commission estimates 600,000 jobs could be eliminated.
Read article
OpenAI and Microsoft have ended their exclusive cloud deal, with GPT-5.5, Codex, and Managed Agents now available on Amazon Bedrock — ending the era where the world's most-used AI was locked to a single cloud.
Read article
Anthropic ran a controlled marketplace where Claude agents autonomously negotiated 186 deals worth $4,000. Frontier models got objectively better outcomes — and the people with weaker agents never noticed.
Read article
OpenAI released GPT-5.5 alongside workspace agents that automate lead qualification, reporting, and IT triage — marking a concrete shift from AI-as-chatbot to AI-as-colleague for business teams.
Read article
Microsoft's A$25 billion investment will expand Azure AI capacity by 140%, train three million Australians in AI skills, and deepen national cyber defence — making it the largest single corporate commitment to Australia's digital infrastructure.
Read article
Thomson Reuters finds organisation-wide GenAI adoption has doubled to 40% in professional services, yet only 18% track ROI — because AI efficiency directly threatens the billable hour model that underpins most firms' revenue.
Read article
OpenAI has moved ChatGPT advertising from CPM to cost-per-click pricing with $3-$5 bids, targeting $2.5 billion in ad revenue for 2026 as it competes directly with Google and Meta for performance budgets.
Read article
A landmark CSIRO study of 4,000+ Australian firms found that companies adopting AI posted 36% more job ads than non-adopters — and professional roles like lawyers and accountants are thriving, not shrinking.
Read article
Salesforce unveiled Headless 360 at TDX 2026, exposing every CRM capability as an API, MCP server, or CLI command so AI agents can operate the platform without a browser. It's the clearest signal yet that the era of clicking through business software is ending.
Read article
Adobe's new Firefly AI Assistant orchestrates roughly 100 tools across six Creative Cloud apps from a conversational interface — turning natural language into multi-step creative workflows.
Read article
Softr, C3 AI, and SnapApp all shipped AI-powered app builders in the same month — signalling that non-technical teams can now build real business software without writing code.
Read article
In March 2026, TIDY launched full-service AI property management at 3.9% of revenue, Entrata deployed 100+ AI agents, and Braiin entered the market — signalling a structural collapse in what landlords pay for professional management.
Read article
Zip launched AI Contract Orchestration and 50+ purpose-built procurement agents, cutting contract cycle times by 55% and outside legal hours in half. Here's what it means for any business that signs supplier agreements.
Read article
KPMG's Global AI Pulse survey of 2,110 C-suite leaders finds Australian businesses lead on AI governance (31% vs 26% globally) but rank last on AI-driven productivity gains (35% vs 42%). With multifactor productivity already declining, the governance-first approach is becoming a strategic gamble.
Read article
Chrome's new Skills feature lets anyone save Gemini AI prompts as reusable one-click workflows across any webpage, bringing no-code AI automation to the 65% of desktop users already running Google's browser.
Read article
OWASP's first Top 10 for Agentic AI, a McKinsey red-team breach in under two hours, and Bessemer's security framework converge on a single warning: autonomous AI agents are 2026's defining cybersecurity threat.
Read article
MYOB and Microsoft have announced a five-year partnership to embed agentic AI directly into MYOB's products for 3.28 million ANZ small businesses, including Australia's first AI-powered BAS preparation tool.
Read article
Major retailers are publishing hard ROI data from AI shopping assistants — Macy's Gemini-powered chatbot drives 4.75x higher spending while Frasers Group lifts conversions 25%, giving business owners the first concrete benchmarks for conversational commerce.
Read article
OpenAI published a 13-page industrial policy paper proposing four-day workweeks, automation-linked taxes, a public AI wealth fund, and a 'Right to AI' — here's what it means for Australian businesses navigating the AI transition.
Read article
Indeed Hiring Lab's April 2026 report reveals that AI adoption in Australian business has doubled in a year — but two-thirds of AI-related job postings still come from just 1% of employers, exposing a widening gap between AI-ready firms and everyone else.
Read article
Anthropic unveiled Claude Mythos Preview through Project Glasswing, a closed cybersecurity initiative with Apple, Microsoft, and Google that has already discovered thousands of unknown software vulnerabilities hiding in the world's most critical systems.
Read article
Anthropic disclosed a $30 billion revenue run rate — tripling in four months — and signed a deal with Google and Broadcom for 3.5 gigawatts of next-generation TPU capacity through 2031.
Read article
OpenAI closed a record $122 billion round at an $852 billion valuation, backed by Amazon, Nvidia, and SoftBank. Here's what the biggest private raise in history signals for businesses adopting AI.
Read article
A landmark BCG and NYU report finds fewer than 10% of hospitality companies generate real value from AI, even as 65% of hotels face staffing shortages and labour costs surge. The early movers are already pulling ahead.
Read article
Anthropic CEO Dario Amodei signed a Memorandum of Understanding with the Australian government on AI safety research, with AUD$3 million in research credits to four institutions — the first deal under Australia's National AI Plan.
Read article
Anthropic just paid $400 million in stock for a 10-person biotech startup. The deal isn't really about healthcare — it's about the coming wave of industry-specific AI tools that will reshape how every business operates.
Read article
New Privacy Act amendments force Australian businesses to disclose how they use AI in decisions about people. The deadline is December 10, 2026. Penalties reach $50 million. Here's what's actually required.
Read article
Google DeepMind released Gemma 4 under Apache 2.0 — its most capable open models, now fully free for commercial use. Running on a single consumer GPU, these models challenge the assumption that powerful AI requires expensive subscriptions.
Read article
Microsoft released three in-house MAI models for transcription, voice, and image generation, signalling a strategic break from its $13 billion OpenAI partnership and driving enterprise AI prices down across the board.
Read article
monday.com launched a platform where you literally post roles and hire AI agents. Hexaware has 600+ ready to deploy. Gartner says 40% of enterprise apps will embed agents by year's end. Here's what this means for your business.
Read article
The cost of AI has dropped 90% in 18 months. McKinsey data shows 72% of companies have adopted it. Here's what that means for Australian operators who haven't started yet.
Read article