Daily brief at 7am Melbourne. Unsubscribe any time.

Friday 19 June 2026

Accenture's $4.1 Billion OT Security Bet Is the Biggest Signal Yet That Critical Infrastructure Is the Next Cyber Frontier

Accenture drops $4.1 billion to own the OT security market, the 'Popa' botnet turns TV boxes into a proxy crime empire linked to a Nasdaq-listed firm, and cybercrime now accounts for a third of all crime across Asia and the Pacific.

Lead story

Accenture's $4.1 Billion OT Security Bet Is the Biggest Signal Yet That Critical Infrastructure Is the Next Cyber Frontier

Consulting giant Accenture has announced it will acquire a majority stake in industrial cybersecurity firm Dragos — valuing it at $3.25 billion — alongside outright purchases of asset-discovery platform runZero and firmware-analysis startup NetRise. The total bill: roughly $4.1 billion. It is one of the largest single moves into operational technology (OT) security ever made by a professional services firm.

To understand why this matters, you need to understand what OT security actually is. Unlike IT security — which protects laptops, databases, and cloud servers — OT security protects the computers that run physical things: power grids, water treatment plants, oil pipelines, manufacturing lines, port logistics. These systems were never designed with the internet in mind, they are notoriously hard to patch, and they are increasingly connected to corporate networks that attackers already know how to navigate.

Dragos is arguably the defining name in this space. Founded by former NSA threat hunters, the company built its reputation tracking nation-state groups targeting industrial systems — groups like Sandworm and Volt Typhoon. Adding runZero (which specialises in mapping everything on a network, including the devices that shouldn't be there) and NetRise (which digs into firmware to find vulnerabilities in the chips and code inside industrial gear) turns Accenture's OT offering into a vertically integrated stack: see the assets, understand the firmware, detect the threats.

The strategic logic is straightforward. Accenture already has deep relationships with the utilities, defence contractors, and manufacturers that run OT infrastructure. It can now walk into those clients with a complete industrial security product suite, not just advisory services. For Dragos, the deal provides distribution at a scale no independent security vendor could build organically.

The timing is not coincidental. Attacks on critical infrastructure have intensified across every region. The US, UK, Australia, and their Five Eyes partners have published a string of joint advisories in the past two years warning that Chinese and Russian state actors are pre-positioning inside OT networks — not necessarily to cause immediate disruption, but to have options in a future crisis. That threat environment is turning OT security from a niche specialism into a board-level procurement priority.

For Australia, the implications are direct. The Security of Critical Infrastructure (SOCI) Act already mandates that operators of critical infrastructure — including energy, water, ports, and communications — manage cyber risk across their assets and supply chains. Australian entities using Dragos or planning OT uplift programs should be watching how the Accenture integration unfolds: if the combined entity lifts prices or shifts product focus toward larger US federal clients, mid-tier Australian operators may find their options narrowing.

What to watch: Whether Dragos's threat-intelligence community culture survives absorption into a 700,000-person consulting machine. Some of the most respected OT researchers in the world work at Dragos; retention will be the acid test of whether this is a genuine capability acquisition or an expensive logo purchase. Accenture has done this kind of deal before — it bought iDefense and Context IS over the years — and the track record is mixed.

The deal is expected to close later this year, subject to regulatory approval. Given Dragos's US government relationships, a national-security review is plausible.

Also today

'Popa' Botnet Secretly Tied to Nasdaq-Listed Israeli Firm

Researchers have connected a four-year-old Android botnet called Popa — which hijacked millions of consumer TV set-top boxes to relay traffic for ad fraud, account takeovers, and large-scale data scraping — to NetNut, a 'residential proxy' service run by publicly-traded Israeli company Alarum Technologies (NASDAQ: ALAR). The allegation is significant: residential proxy networks are sold as legitimate tools, but this research suggests some of that traffic flows through devices whose owners have no idea they're part of the network. It raises uncomfortable questions about how proxy providers source their IP addresses — and who is ultimately responsible when those sources turn out to be compromised consumer devices.

Krebs on Security

Cybercrime Is Now a Third of All Crime Across Asia and the Pacific

Interpol's latest regional review finds that cyber offences now account for approximately one in three crimes recorded across Asia and the South Pacific — a figure that would have been unthinkable a decade ago. Scams continue to dominate the category, and AI-enabled attackers are outpacing the response capacity of cash-strapped regional law enforcement agencies. The report underscores a growing enforcement gap: the technical sophistication of attacks is rising faster than governments in the region can train investigators or fund defences. For Australia, which sits at the wealthier end of the regional spectrum, the finding reinforces the case for the kind of upstream blocking measures flagged in recent ACSC guidance.

The Register

DragonForce Hides Ransomware C2 Inside Microsoft Teams Traffic

Researchers at Symantec and Carbon Black have caught the DragonForce ransomware group using a custom Go-based remote access trojan — named Backdoor.Turn — that tunnels its command-and-control traffic through Microsoft Teams relay infrastructure. By piggybacking on a legitimate, widely-trusted enterprise platform, the malware blends into the kind of traffic that most organisations explicitly whitelist. The technique is a significant operational security upgrade for a group that has already compromised major targets. Security teams relying on network-layer blocking to detect C2 channels need to account for the growing use of legitimate SaaS platforms as camouflage — a trend that shows no sign of slowing.

The Hacker News

F5 Patches Two Critical RCE Flaws in NGINX

F5 has pushed out-of-band patches for a pair of critical vulnerabilities in NGINX Open Source, including CVE-2026-42530 (CVSS v4: 9.2), a use-after-free bug in the HTTP/3 module that a remote, unauthenticated attacker could trigger to execute arbitrary code. The second flaw is similarly severe. NGINX is the world's most widely deployed web server, underpinning a vast proportion of internet infrastructure including many Australian government and commercial web properties. The out-of-band timing signals F5 considered the risk serious enough not to wait for its regular patch cycle. Administrators running NGINX with HTTP/3 enabled should treat this as urgent.

SecurityWeek

ASD Tightens Security Skills Requirements for Developers

Australia's Australian Signals Directorate has updated its Information Security Manual with new and revised controls that draw a harder line on developers who lack baseline security skills. The updated ISM signals that security-by-design is no longer a nice-to-have for government-adjacent projects — it is an explicit compliance requirement. The move aligns with the broader push from ASD and the Department of Home Affairs toward secure-by-default software development practices. Organisations building or procuring software for government use should review the updated controls, particularly around secure coding standards and developer training obligations.

iTnews

Icarus Group Exploits Klue OAuth Flaw to Steal Salesforce CRM Data

A threat actor tracked as 'Icarus' has compromised market intelligence platform Klue by exploiting an OAuth misconfiguration, using that access to pivot into the Salesforce CRM instances of multiple downstream organisations in an ongoing extortion campaign. Cybersecurity vendor Huntress is among the confirmed victims. Klue's Battlecards product is now the third Salesforce-integrated application to be weaponised this way in recent months, suggesting Icarus has developed a repeatable playbook: find a poorly-secured OAuth integration, inherit the permissions of the connected CRM, and exfiltrate the data. Any organisation with third-party apps connected to Salesforce via OAuth should audit what access those apps actually hold.

Bleeping Computer

Malware Developers Are Stuffing Weapons-Talk Into Code to Fool AI Scanners

At least one malware author has started embedding fake policy-triggering text — references to nuclear and biological weapons — inside JavaScript comment blocks, with the sole purpose of causing AI-powered analysis tools to refuse to process the file. The actual malicious code sits below the comment and runs normally; the content-safety guardrails of AI analysis platforms never see it. It is a clever, low-cost evasion technique that exploits the same content policies that make AI tools safe for general use. The finding has implications for security teams that rely on AI-assisted triage: the adversaries have already started to map where those tools flinch.

Schneier on Security

OpenAI Recruits Transformer Co-Inventor and Trump's AI Policy Chief Ahead of IPO

OpenAI has landed two high-profile hires in the same week as it prepares for its public listing: Noam Shazeer, co-inventor of the Transformer architecture that underpins virtually every modern large language model (he came from Google DeepMind), and Dean Ball, who shaped AI policy inside the Trump administration. The pairing is deliberate — Shazeer adds research credibility and technical firepower, while Ball signals to regulators and Washington that OpenAI can work within the political mainstream. Both hires are clearly timed to shore up investor confidence before the IPO. The company is trying to look simultaneously like the most capable AI lab and the most governable one.

TechCrunch

Amazon Moves to Sell Its Trainium Chips to Rival Data Centres

AWS is in active discussions to sell its custom Trainium AI chips to other data centre operators — not just use them internally — in a direct challenge to Nvidia's dominance of the AI accelerator market. CEO Andy Jassy has framed this as a $50 billion revenue opportunity. The move is strategically significant: Amazon has until now treated its custom silicon primarily as a cost-reduction lever for its own cloud. Selling chips externally means competing on hardware, not just cloud services. Whether Trainium can win on performance-per-dollar against Nvidia's H100 and B200 chips — which have near-religious adoption among AI model trainers — remains the central question.

TechCrunch

Novo Nordisk GitHub Token Leak Exposes Dev Pipeline Risk

A leaked GitHub personal access token at pharmaceutical giant Novo Nordisk has exposed a vulnerability pattern that security researchers say most organisations get fundamentally wrong: treating secrets management as a tooling problem rather than an identity governance problem. The token, had it been exploited more broadly, could have granted access to the company's software development pipeline. The incident is a reminder that developer credentials — API keys, tokens, and service account secrets — are often the softest entry point into an otherwise well-defended organisation, and that rotating secrets after discovery is only half the solution if the underlying access model hasn't changed.

Dark Reading

US Congress Revives the No FAKES Act to Target AI Deepfakes

US lawmakers have reintroduced the No FAKES Act, which would create a federal right to control AI-generated likenesses and voice replicas, targeting the use of deepfakes that profit from an artist or performer's identity without consent. The bill has bipartisan support in principle, but faces opposition from digital rights groups concerned about overbroad liability for platform hosts. The legislation is part of a broader wave of AI identity regulation — the EU's AI Act already covers some of this ground, and several US states have passed their own versions. Australia currently lacks equivalent federal legislation, though the Online Safety Act and existing personality rights under state law provide partial coverage.

CyberScoop

Sources consulted