Thoughtworks, a Chicago-based technology consultancy, is shifting its guidance from what artificial intelligence (AI) agents can do to how companies keep them secure, supervised and affordable as the tools take on larger, longer-running tasks.

In volume 35 of its Technology Radar, published on Wednesday, the consultancy said the hardest engineering challenges increasingly lie beyond interactions with the model itself. They include securing the environments agents run in, deciding what they can access, keeping human oversight meaningful and controlling running costs.

“Not long ago, the question we heard from clients was how fast agents could write code. Now it’s whether anyone can trust, review and afford what they produce,” said Rachel Laycock, chief technology officer at Thoughtworks. “Models will keep improving, but lasting value will go to the organizations that invest in the engineering and oversight around them.”

The consultancy, which integrates design, engineering and AI, publishes the Radar twice a year based on its work with clients.

The latest edition places 11 practices and technologies in its Adopt ring, the Radar’s strongest recommendation. They include agent sandboxing and threat modeling, which the report said signal the growing importance of safeguards for running AI reliably in production.

Systems around agents

The report said the “harness” around an AI agent, including its prompts, tools, memory and guardrails, is becoming a product in its own right, with owners, versions and dependencies. As plugins multiply, it said companies must treat it as an engineered system, with evaluation, security, access controls and maintenance.

Developers increasingly run several agent sessions in parallel and hand off longer tasks, including unattended overnight builds.

But as agents produce more work, human review and software delivery pipelines can become bottlenecks. The Radar warned against “human in the loop theater,” where approval becomes a ritual rather than a meaningful check, and urged teams to assess their review capacity before scaling up agent use.

It also said proprietary frontier models are no longer the obvious default, as open-weight models have narrowed the gap for many workloads and can run in-house. It urged teams to weigh cost and privacy when choosing models.

The cost of running AI also extends well beyond model pricing.

Routing, batching, storage, networking and scheduling all shape the economics of production workloads. The report said much of the innovation needed to make AI economically sustainable lies deeper in the stack, where models, graphics processing units (GPUs), storage and networks are managed as programmable resources through a common platform layer.

Thoughtworks concluded that realizing the potential of enterprise AI will take more than access to capable models. Companies must build the engineering practices, infrastructure and human oversight needed to operate agents responsibly at scale, it said.

Elastic, a San Francisco-based search and security software company, is warning that attackers are running commands through artificial intelligence (AI) agents, abusing approved information technology (IT) tools and logging in with stolen passwords within minutes, turning the trust enterprises place in such software against them.

“What’s new is that enterprises are increasingly deploying powerful trusted tools that most security teams struggle to fully monitor, control or even explain: AI agents. An agent that holds credentials and can reach the internet is a new attack surface, whether it’s misconfigured or abused,” Mike Nichols, general manager of security at Elastic, said in a press release.

“When an attacker is behind it, its activity looks exactly like the employee who installed it. So the question for defenders isn’t whether a tool is approved; it’s what that tool is actually doing.”

Elastic sells search, observability and security software built on its Elasticsearch engine.

The findings come from threats, etc, a quarterly report series from Elastic Security Threat Command drawing on confirmed intrusions and research between April and September 2026. The company said it publishes quarterly because attacker tradecraft, including the abuse of AI agents, now changes too fast for an annual cycle.

In live intrusions, Elastic researchers saw attackers execute commands through Claude Code on Windows and the ChatGPT desktop app on macOS. In two of those cases:

• On a Mac, a single command read an internal infrastructure document and then pulled saved passwords from the user’s keychain

• On Windows, an attacker connected remotely, used Claude Code to stage tools and reached an outside server through a concealed connection

Elastic also forensically reconstructed the July breach at Hugging Face, an AI model-sharing platform, where an autonomous AI agent broke into production systems. Signs of the agent at work included duplicate commands seconds apart and a working exploit followed immediately by a basic coding error.

Stolen passwords at speed

Across 27 confirmed cases involving password-stealing malware, attackers logged in with stolen credentials an average of eight minutes after infection.

The most common entry point was a “ClickFix” lure, a fake prompt that tricks users into pasting and running a malicious command themselves. It started in 12 of the 27 cases, or 44%. The theft then kept running inside software the device already trusted, such as browser updaters and Apple’s built-in scripting tools.

In one healthcare case, attackers delivered malware about 16 seconds after tricking the user.

The company also confirmed 23 cases in which attackers installed legitimate remote IT software, most often ScreenConnect, to keep hands-on access to a device.

Elastic recommends that defenders inventory the AI tools, IT tools and signed software they trust, and judge them on behavior, such as what an AI tool launches or where a remote IT tool connects, rather than on whether they are approved.

PagerDuty, a San Francisco-based incident management company, has made its site reliability engineering (SRE) agent, Paige, generally available as more software written by artificial intelligence (AI) reaches production with less human review.

Once live, AI-powered software also fails in new ways, from model drift and runaway token use to hallucinations and exposure of personal data. As more applications add features powered by large language models (LLMs), outages are becoming more frequent and the teams keeping business-critical systems running are under growing pressure.

“A tsunami of fragile AI-generated code is now entering production and SRE teams are stretched to the max. This new code breaks differently and more often and practitioners find it increasingly difficult to determine the root cause when a problem does arise,” said John DiLullo, chief executive of PagerDuty, in a press release.

“The only way to maintain resilience in the face of this complexity is to fight fire with fire, using AI itself as part of how we keep systems running.”

PagerDuty manages close to one billion incidents a year and is used by about two-thirds of the Fortune 100.

Humans in the loop

Teams can keep a human in the loop, letting Paige speed up triage while holding back remediation for approval, or let it act on its own in well-understood scenarios. It automatically resolves lower-priority incidents while giving human operators full visibility into what it does.

Each resolved incident is added to a shared memory that the agent draws on in later cases.

Paige runs on the PD Reliability Platform, the renamed PagerDuty Operations Cloud, which works inside Slack and Microsoft Teams. When an incident triggers, the platform opens a dedicated channel and conference bridge, then drafts a post-incident review and assigns follow-up actions to owners.

PagerDuty unveiled the agent as a virtual responder at its London event in March and opened early access in the second quarter.

Snow Tempest, research manager for information technology (IT) service management at IDC, said IT teams want measurable returns from AI agents, including better uptime, lower operating costs and less time spent chasing logs, and that organizations increasingly aim to prevent incidents as well as recover and learn from them.

Kulvir Gahunia, director of the site reliability office at Canadian telecom group TELUS, said an AI scribe agent from PagerDuty runs through every incident in its internal tool, so responders do not have to stop and document what is happening.

Alongside the launch, PagerDuty is opening PD Labs, an innovation center where design partners can try early tools such as SRE Workbench on real problems and work directly with the engineers writing the code, as AI changes faster than any traditional product roadmap can keep up with.

Cognition, a San Francisco-based artificial intelligence (AI) lab, is expanding in London to serve UK enterprises adopting AI software engineering.

Its UK team has grown by more than 200% in six months, and the London hub will support operations across Europe, the Middle East and Africa (EMEA), the company said.

“The UK is one of the world’s most important technology and financial services markets and we’re seeing significant demand from organizations ready to put AI software engineering to work,” said Russell Kaplan, co-founder and president at Cognition. “This is a long-term investment for Cognition. We want London to become a major hub for our growth across the UK and EMEA.”

Founded in 2023, Cognition developed Devin, an autonomous AI agent that plans, writes, tests and deploys code.

Narry Singh has joined as Senior Managing Director. He most recently served as global co-head of digital at AlixPartners and previously as global head of growth and strategy at Accenture Digital. He is also a former non-executive director of the London Stock Exchange.

Cognition said Singh will work with chief executives, finance chiefs, and boards to build the business case for AI software engineering, put it to use responsibly across enterprises, and manage the workforce and operating-model changes that follow.

Aman Thind has joined as field chief technology officer (CTO) for financial services. He was previously managing director and global head of technology strategy and architecture at JPMorgan Chase, and chief architect at State Street, where he led the firm’s AI, cloud and digital asset strategy.

He also co-founded BestX, a fintech State Street acquired in 2018.

Thind will partner with chief information officers, CTOs and engineering leaders to bring AI agents into software development and support Cognition’s banking and financial services customers globally, the company said.

In the UK, The Access Group, a business software provider, has deployed Devin to more than 2,000 engineers. Cognition said the rollout contributed to a fivefold increase in merged pull requests and sped up some legacy software migration tasks by up to 480 times.

Across Europe, Cognition recently partnered with McKinsey & Company to help enterprises scale AI-powered software development and get more return from their AI spending.

Rival coding platform Cursor announced a London headquarters in June, saying its EMEA revenue had tripled quarter over quarter.

Cognition is widening how enterprises can access its tools. In September, it signed a multi-year collaboration with Amazon Web Services (AWS) to help companies speed up cloud modernization, clear security and engineering backlogs, and run autonomous software engineering inside existing AWS environments.

Backed by a September round that raised more than $2 billion at a $48 billion valuation, Cognition plans to keep growing its London team as it chases enterprise demand.

Cisco, a San Jose-based networking and security company, warns that the rush to deploy AI agents is outpacing companies’ ability to monitor them, and unpredictable bills are now a bigger headache than high ones. The findings come from its Splunk State of Observability 2026 report, released on October 20.

Oxford Economics surveyed 2,050 IT operations (ITOps) and engineering professionals in 10 countries, including the UK, between February and March.

“Most companies will soon have far more agents at work than people,” said Jeetu Patel, Cisco's president and chief product officer. “Managing agents at scale requires reimagining both security and observability.”

The report found that 55% of respondents could not accurately control their AI costs.

An additional 60% said unexpected billing spikes caused more operational strain than high baseline costs, while 57% said telemetry costs had overtaken the infrastructure they were monitoring.

“For today’s enterprises, the excitement of AI is clashing with the reality of the bill,” said Kamal Hathi, senior vice president and general manager of Splunk at Cisco. “We believe observability is the essential bridge to restoring predictability and confidence to the agentic AI era.”

Concern is widespread, with 83% of respondents worried about AI agents' unintended consequences.

While 87% said monitoring AI was a priority, 54% had no clear business outcomes tied to it. More than half (56%) could not measure metrics such as large language model (LLM) response quality, and while 85% said visibility into hallucinations was important, only 58% said they tracked them effectively.

Monitoring AI has also pulled IT and security teams closer, with 81% saying it had increased collaboration. Joint troubleshooting rose to 77% from 62% in 2025.

“This report shows that observability and security data are converging, and that teams need to collaborate more than ever,” said Patel. “Common data platforms that help us detect and respond to issues in real-time are the key to deploying agents we can trust at scale.”

Some 32% of UK respondents said they used AI agents all the time to diagnose root causes, compared with 20% globally. At the same time, 55% said they were strongly worried about agents' unintended consequences.

Elsewhere, 51% of Indian firms said their monitoring of AI model quality was very effective, compared with 35% globally. In Saudi Arabia, 98% said AI let them focus more on innovation, while 74% of German respondents rated visibility into AI privacy as very important.

The report identified a group of leaders who closely monitor model quality, cost and security. Of these, 91% could accurately attribute AI costs to specific agents and models, compared with 69% of other firms.

First Light Fusion, an Oxford-based fusion developer, will fire projectiles into graphite at over 3 km/s to help UK defense scientists study extreme impacts.

The tests will study crater formation, energy dissipation and transitions in material strength.

“This program demonstrates the growing demand for First Light Fusion’s capabilities beyond fusion energy,” said Paul Holligan, vice president of commercial and partnerships at First Light Fusion.

Founded in 2011, the company is developing FLARE, a two-stage fusion fuel system.

In March 2025, it dropped plans to build its own fusion power plant, turning to partnerships with other fusion developers and to materials testing for space and defense. Six months later, it published a paper setting out FLARE as a reactor-compatible route to high-gain inertial fusion.

The campaign for the Defence Science and Technology Laboratory (Dstl), the research arm of the UK Ministry of Defence, will reproduce a 2017 experiment by Hébert and colleagues that fired steel projectiles into porous graphite.

Replicating the study allows direct comparison with published results and tests the accuracy of First Light’s methods and measurements.

The work will also probe a change in graphite’s behavior observed at impact velocities above about 3 km/s. Dstl will supply the targets, which are configured for recovery after each shot so researchers can inspect the damage in detail.

The laboratory said the program would deepen its understanding of how materials behave under hypervelocity impact. It said First Light’s facilities and expertise would let it examine these effects across a carefully controlled range of conditions, supporting the development and validation of improved material models for defense applications.

First Light’s gas gun and pulsed-power platforms, built for its fusion work, can recreate the extreme temperatures and pressures relevant to aerospace, defense and space testing. They include the UK’s largest two-stage gas gun.

In June, its VIPER amplifiers pushed projectiles past 12 km/s, comparable to orbital debris, at a Texas A&M test station.

First Light expects demand for its fusion-built facilities, amplification technology and simulation tools to keep growing among government and commercial organizations in defense, aerospace and advanced materials research.

Large enterprises report spending an average of 28% of their software budgets on upgrades, migrations and licensing changes they would not have chosen, a survey shows.

In total, 95% of information technology (IT) leaders surveyed said their organization had paid in the past three years for a vendor-driven upgrade it believed would reduce performance, reliability or business value. The latest such change cost an average of $12.87 million.

“When the alternative to an unwanted upgrade is losing support on systems the business runs on, saying yes is often a rational response. That’s exactly the problem,” said Tomás O’Leary, founder and chief executive of Origina.

“The market has left some of the most experienced technology leaders in the world with little room to refuse, and that cost shows up in budgets, in engineering time and in innovation that never gets funded. We want to make that cost visible so customers can start to take back control of it,” he said.

Origina, a Dublin-based independent software maintenance provider, commissioned the research and published it in a report with Thinkers50.

Censuswide surveyed 2,018 chief information officers (CIOs) and senior IT decision-makers in the US and UK in May at organizations with revenues of at least $1 billion or 10,000 or more employees.

Most said they have little room to push back:

• 78% said their main vendor sets the timing and terms of change

• 76% said they have very little real leverage in negotiations

• 80% said leaving a major vendor would cause significant disruption

Respondents most often named innovation as the budget hit by these costs, with hiring and security close behind.

Asked about the last time an artificial intelligence (AI) or data program was slowed, paused or canceled, 34% said vendor-driven costs had made a major contribution and 56% a moderate one.

Some 91% said they had scrapped functional hardware in the past three years purely to meet a vendor’s support deadline, losing an average of 1.65 years of useful life. Meanwhile, 84% said senior engineers always or often spend time on vendor upgrades rather than building future capability.

James McGlennon, former CIO of Liberty Mutual and a member of Origina’s board, said a single upgrade cycle at the insurer’s scale could occupy some of its strongest technical people for most of a year, a cost that recurs every year.

More than nine in 10 respondents (91%) would back legislation giving software customers protections like those of long-term residential tenants.

Marshall Van Alstyne, a Boston University professor who co-wrote the report with O’Leary, said customers who make relationship-specific investments hand vendors leverage no contract can fully anticipate. He said CIOs should negotiate the exit before the entry, because the ability to leave has value even if they don't leave.

The report sets out steps to reduce lock-in, from modular software architecture to contracts that settle foreseeable decisions before customers commit.

Akamai Technologies, a Cambridge, Massachusetts-based cloud and cybersecurity company, announced on September 24 that it signed a seven-year, $11.6 billion agreement to support Anthropic's growing central processing unit (CPU) workloads, the developer of the Claude artificial intelligence (AI) models.

The commitment could rise by a further $9 billion, lifting the relationship’s potential value to about $20 billion.

Anthropic will tap Akamai Cloud’s distributed AI infrastructure and software to meet its accelerating CPU workload demands. The contract adds to more than $2.8 billion in multi-year cloud infrastructure services (CIS) commitments that Akamai announced earlier this year across its customer base.

Tom Leighton, co-founder and chief executive of Akamai, said Anthropic was advancing the AI revolution and had chosen his company’s capabilities to build and operate AI infrastructure at scale.

“Our expanding global footprint, combined with our years of experience serving the world’s largest enterprises, positions us to be the infrastructure provider for secure and responsible AI applications and workloads,” he said.

Akamai Cloud supports computing from the core to the edge on a network spanning thousands of points of presence. Built on diversified hardware, it lets customers build and run applications and optimize how they serve users and AI agents.

It is designed to keep applications fast, reliable and secure across their full lifecycle, as customer demand for deploying AI workloads grows.

Akamai has also given Anthropic the right to buy up to about 5% of its shares at $111.33 each. About 2% is tied to the initial commitment. Anthropic can unlock the remaining 3% in 1% steps for every additional $3 billion it spends on Akamai’s cloud services over seven years.

Akamai estimates capital expenditure tied to the contract at about $5.5 billion. It expects no impact on its 2026 revenue guidance but plans to raise this year’s capital spending by about $1.7 billion to secure and pre-purchase critical supply chain components, including memory.

Guarding nonhuman identities

The deal came two days after Akamai published its latest State of the Internet (SOTI) security report on September 22, which warns that agentic AI is fundamentally changing enterprise risk.

Titled The Agentic Frontier: Speed, Scale, and Nonhuman Identity and aimed at chief information security officers (CISOs), the report said securing a business has evolved beyond managing people's identities and access to behavioral governance over nonhuman entities that execute tasks across application programming interfaces (APIs) and systems.

It identified two major challenges for CISOs. One is setting guardrails for autonomous agents. The other is how quickly AI models discover and chain vulnerabilities, often outpacing human patch cycles.

Key findings include:

• Model Context Protocol (MCP) exposure ranks last among CISO security priorities, even though the protocol lets AI agents act autonomously across multiple software systems and security leaders expect rogue agents to become a top cyberthreat by 2030

• More than 6% of enterprise AI chatbot conversations contain sensitive corporate data, mainly personally identifiable information, with 47% of interactions taking place via unmonitored personal accounts

• More than 40% of enterprise users have installed AI-powered browser extensions, a quarter of which changed their permissions within 12 months; these tools are 60% more likely than standard extensions to carry known common vulnerabilities and exposures (CVEs)

• AI bot traffic rose 300% in 2025, with the commerce industry hit hardest

• The average annual cost to an enterprise of resolving API breaches has reached $700,000

Boaz Gelbord, chief security officer at Akamai, said companies face a triple threat from internal AI use, such as AI-generated code and productivity tools; AI built into customer-facing products and cloud workloads; and AI-driven attacks on the enterprise.

“AI presents an ‘everything, everywhere, all at once’ moment for the security ecosystem as a whole. Security programs need to adapt to this rapidly evolving reality, and there’s a lot of pressure in the system due to the unprecedented speed of these changes,” he said.

The report urges CISOs to:

• Deploy edge-native runtime protections, API filters and isolation to neutralize vulnerabilities while back-end patching takes place

• Build visibility and behavioral controls inside the browser to secure staff use of unmanaged AI extensions and web-hosted models

• Use generative engine optimization (GEO) and machine-readable data at the network edge to stop AI crawlers and synthetic shoppers misrepresenting brand data

• Grant AI agents autonomy according to how easily their actions can be verified and how reversible a failure would be, keeping humans in the loop for high-stakes actions

Gelbord said the issue will be a central topic for boards, customers and regulators for the foreseeable future.

AI agents are gaining direct, governed access to the systems that already run corporate sales operations, a shift from assistants that merely suggest actions to ones that can execute them inside existing business rules.

Salesforce and Anthropic announced on August 26 Claudeforce, an expanded partnership that connects Claude’s reasoning with the data, workflows, business logic, actions, and governance of the Salesforce platform.

The partnership launches with Salesforce in Claude, a plugin with 37 prebuilt sales skills covering meeting preparation, deal health review and pipeline review. Sellers can reason over live revenue context, automate pipeline updates and take governed action from within Claude.

An onboarding process reads a seller’s enterprise context from Salesforce, Slack, and other connectors to build a tailored dashboard of accounts and pipeline data, with interactive views generated by Claude.

Salesforce in Claude routes actions back through Salesforce, ensuring each company’s business rules are enforced when an action is taken. An administrator connects the plugin once, with authentication and permissions centrally managed for the entire sales team.

Salesforce in Claude runs on AIforce, Salesforce’s enterprise harness that exposes company data and workflows to agents through Model Context Protocol (MCP) servers, application programming interfaces (APIs) and command-line interface (CLI) tools.

“Salesforce in Claude brings this same frontier intelligence into the systems where much of the world’s commercial activity happens,” said Dario Amodei, chief executive and co-founder of Anthropic. “Companies can point Claude at the customer information and business context that they’ve been building in Salesforce for decades, and use it to actually run and grow their businesses.”

Marc Benioff, chair and chief executive of Salesforce, said the partnership fuses Claude’s reasoning with the trusted data, workflows and governance that enterprises already run on, aiming for a system that can think, reason and act rather than one that only automates fixed tasks.

The integration runs in both directions. Claude now serves as a reasoning model for Salesforce’s Atlas Reasoning Engine, powers Agentforce Vibes and Agentforce Coworker by default and is available in Agent Builder. Through Amazon Bedrock, Claude also operates within the Salesforce Trust Boundary, allowing regulated customers to deploy it while keeping data and AI workloads secure.

Claude is also becoming the default model across Slack, powering Slackbot, augmenting team decisions through Claude Tag and serving as a founding partner for Slack Code. Internally at Salesforce, Slackbot running on Claude has driven 8.1 million annualized hours of productivity gains, more than double the prior quarter.

Salesforce in Claude is now available to select pilot customers and is expected to enter open beta in September 2026, with additional prebuilt skills arriving in the third quarter of 2026.

Schneider Electric has released a study concluding that arc flash risk in 800-volt direct current (VDC) data center power systems can be managed and, in many cases, matches the safety profile of conventional alternating current (AC) systems, the company said.

The French energy technology firm supplies power distribution, industrial automation and data center equipment and software, including the ETAP power system simulation platform it majority owns.

The findings arrive as 800 VDC architectures gain adoption in AI data centers to support denser server racks beyond 400 kilowatts, a shift driven partly by NVIDIA’s push for higher power AI computing hardware. Unlike AC systems, converter-fed 800 VDC systems currently lack an industry-wide standard for managing arc flash hazards.

The study used ETAP’s simulation software and digital twins to model fault scenarios.

Two 800 VDC architectures were evaluated: a rack-level “sidecar” design and a centralized, facility-level design, both reflecting deployment patterns already used by major hyperscale operators.

The rack-level case study found incident energy remained well below the referenced 1.2 cal/cm² threshold for personal protective equipment (PPE), even without protection devices installed. The facility level design showed slightly higher incident energy under conservative assumptions with no overcurrent protection, though adding standard protection devices brought energy in line with typical AC architectures.

Capacitor discharge dominates the first milliseconds of a fault event, and standard assessment methods tend to overestimate risk in capacitor-dominated systems, the study found.

“800 VDC power distribution represents a significant shift in data center design, but it also introduces safety considerations that need to be studied extensively,” said Manish Kumar, executive vice president of secure power and data centers at Schneider Electric. “Our work with major hyperscalers provides engineers and safety professionals with one of the first practical frameworks for evaluating arc flash risks.”

“Industry standards remain essential for arc flash and electrical safety, but traditional methods can be overly conservative because they do not fully reflect how complex DC systems operate,” said Tanuj Khandelwal, chief executive of ETAP.

The study also found that fault locations upstream and downstream of reverse blocking diodes affected back-feed and peak current, shaping facility-level outcomes. Schneider Electric said the work builds on its history of arc flash safety testing, including validation of live swap power capabilities in 800 VDC systems.

Online retail has become the world’s most heavily targeted industry by cybercriminals, driven by a rapid shift toward agentic commerce, a model in which autonomous artificial intelligence (AI) agents browse, compare and purchase on behalf of human users. As that shift accelerates, it is creating a new class of security vulnerability that traditional defenses are failing to catch.

By December 2025, nearly half of all commerce traffic across Akamai’s global network consisted of AI bots. Commerce accounted for 48% of all AI bot activity across every industry the firm monitors, more than three times the share of the second most targeted sector.

“We are securing a digital frontier where the ‘customer’ is increasingly an AI agent operating on behalf of the human user,” Patrick Sullivan, chief technology officer of security strategy at Akamai, said in a press release. “This report reveals how and why security leaders must embrace ‘agentic readiness’ to architect sites that welcome legitimate AI while aggressively shutting down malicious bots.”

The findings come from Akamai’s State of the Internet (SOTI) report, Securing the Agentic Storefront: Attacks on Commerce, published July 15, covering January 2024 through December 2025.

The report’s most alarming finding concerns fraud enabled by agentic commerce itself. Because AI shopping agents precisely mimic human browsing behavior, traditional bot detection has been rendered largely obsolete. Threat actors exploit this through agent hijacking, compromising a legitimate AI assistant to abuse its stored payment credentials and execute unauthorised purchases before the customer or retailer notices.

Large language models (LLMs) are also being deployed to generate synthetic identity fraud through “Frankenstein” accounts, fabricated profiles blending genuine personal data with invented elements to bypass standard fraud checks.

Application programming interface (API) exposure compounds the problem. Web attacks targeting APIs rose 9% year over year between the fourth quarters of 2024 and 2025. A companion Akamai API Security Impact Study found that 85% of commerce respondents experienced at least one API-related incident in the past year, yet only 22% know which of their APIs expose sensitive data.

Commerce platforms were targeted by application-layer distributed denial-of-service (DDoS) attacks nearly 3 trillion times in 2025, with retail absorbing 84% of that volume. Attackers use HTTP botnets timed to peak shopping periods, blending into genuine traffic surges to exhaust servers and halt sales.

Between July and December 2025, the retail vertical in EMEA recorded 26 billion AI bot counts, representing 12.4% of all AI bot activity.

EMEA commerce platforms faced 40 billion web attacks between January 2024 and December 2025. Average daily phishing volume rose from 56,600 in February 2026 to 134,600 in April, fuelling account takeover (ATO) attempts and loyalty point theft.

The report urges security leaders to adopt risk-based bot governance, continuously inventory API estates, and deploy microsegmentation to contain the damage from breaches. Only 35% of organisations have achieved true microsegmentation despite 92% applying basic network segmentation.

Akamai is a Cambridge, Massachusetts-based cybersecurity and cloud computing company whose network spans approximately 340,000 servers across 130-plus countries.

Schneider Electric will present its latest artificial intelligence (AI) infrastructure solutions at Datacloud Global Congress next week, including validated AI factory blueprints for NVIDIA’s GB300 NVL72 platform and the Grace Blackwell Ultra architecture.

The energy technology company, which employs 160,000 people across more than 100 countries, will demonstrate liquid cooling systems, 800-volt direct current (VDC) power architectures, digital twin capabilities and end-to-end software tools designed to help operators build and run large-scale AI data centers.

The showcase comes as AI infrastructure investment accelerates sharply. Nearly $3 trillion of AI-related spending is expected to flow through the global economy by 2028, according to Morgan Stanley Research, while Gartner forecasts worldwide AI spending will exceed $2.5 trillion in 2026 alone.

Several Schneider Electric executives will lead keynote panels and technical sessions at the congress, which runs from June 1 to 4 in Cannes, France, and will also co-host an invitation-only executive briefing with NVIDIA on the future of AI-driven infrastructure.

“AI is fundamentally reshaping the future of digital infrastructure, creating new demands around power, cooling and resiliency at unprecedented scale,” said Marc Garner, Global President, Cloud and Service Provider Segment at Schneider Electric. “At Datacloud Global Congress, we will demonstrate how collaboration across the ecosystem is enabling the next generation of AI factories and helping organizations build scalable, resilient and sustainable infrastructure built for the AI era.”

Neoclouds, cooling, 10MW

On June 2, Frédéric Godemel, Executive Vice President of Energy Management Business at Schneider Electric, will join executives from Oracle, DATA4, QTS Data Centers and CBRE for a keynote panel on how the data center ecosystem is keeping up with AI demand, covering the rise of neoclouds, European competitiveness and the gap between hyperscale and enterprise requirements.

Also on June 2, Thierry Chamayou, Vice President of Cloud and Service Providers in EMEA at Schneider Electric, will join a panel with GreenScale, Trench Group, Kao Data, JSM Group and Solar Turbines to explore how operators can de-risk energy investments through smarter project structures and closer collaboration with utilities and government.

On June 1, Sébastien Cruz-Mermy, Vice President of Datacenter Innovation at Schneider Electric, will lead a technical session on AI factory infrastructure, focusing on ultra-high-density rack design, next-generation DC power delivery and resilient cooling. Schneider Electric and NVIDIA will also co-host an invitation-only executive briefing on NVIDIA’s 5-Layer Cake framework and the DSX Blueprint, supported by digital twin tools.

At Stand 122, highlights will include the MCDU-70, a 2.5-megawatt coolant distribution unit (CDU) from Motivair by Schneider Electric that can scale to 10 megawatts and beyond, as well as EcoStruxure IT data center infrastructure management (DCIM) software, NVIDIA Omniverse integrations, and NVIDIA Reference Designs for the GB300 NVL72 platform.

The company’s goal is to help customers design, simulate, build, operate and maintain AI-ready infrastructure at scale, with energy efficiency and sustainability as central requirements.

Schneider Electric, a global energy technology company, and its liquid cooling subsidiary Motivair by Schneider Electric have completed a phased delivery of more than $290 million in AI infrastructure solutions at TeraWulf’s Lake Mariner data campus in Barker, outside Buffalo, New York.

The project converts a legacy industrial site into a series of purpose-built AI and high-performance computing (HPC) data centers within a twelve-month timeframe. Upon full buildout, the 750-megawatt (MW) campus is projected to become one of the largest such deployments underway in the United States, supporting artificial intelligence (AI), high-performance computing (HPC) and cloud workloads for anchor tenants Core42 and Fluidstack, the latter backed by Google.

“TeraWulf’s strategy is centered on delivering scalable, energy-efficient infrastructure capable of supporting the increasing intensity of AI and HPC workloads,” said Sean Farrell, Chief Operating Officer at TeraWulf. “By working closely with industry leaders like Schneider Electric and Motivair, we are accelerating the development of AI-ready capacity at our Lake Mariner facilities while reinforcing the strong operational foundations needed to support long-term customer demand.”

“As demand for AI infrastructure accelerates, ‘time to power’ has become a defining constraint on growth. Operators need partners who can bring together advanced infrastructure, services, and expertise in energy technology to underpin large-scale AI data center deployments at pace,” said Manish Kumar, Executive Vice President, Secure Power and Data Centers at Schneider Electric.

“Our partnership with TeraWulf establishes a strategic blueprint for pairing on-site power, AI-enabled automation, advanced liquid cooling, and digital intelligence at a legacy industrial site. We are delivering resilient, efficient, and scalable data center solutions at the speed and scale this AI era demands,” he added.

Strategically located to leverage low-cost, reliable power, the Lake Mariner campus draws from a New York regional power grid whose energy mix is approximately 89% zero-carbon, with substantial surplus power available. TeraWulf is leveraging its partnerships with Schneider Electric and Motivair to drive long-term growth at speed and scale across the campus.

Liquid cooling at scale

The project required Schneider Electric and Motivair to meet demanding construction and operational timelines and to provide technical design and engineering guidance throughout. The build includes Galaxy VX Uninterruptible Power Supply (UPS) units, Galaxy Lithium-ion Battery Systems and NetShelter Racks and Enclosures for power infrastructure.

On the cooling side, Motivair supplied Coolant Distribution Units (CDUs), In-Rack Manifolds and ChilledDoors. Schneider Electric’s EcoStruxure IT Data Center Expert software was integrated to enable advanced monitoring and digital intelligence, while Motivair’s client services team was deployed to anticipate risks, minimize disruption, and maximize the cooling investment.

Liquid cooling has shifted from a niche technology to a core requirement for modern AI infrastructure, as the thermal output of GPU-dense servers far exceeds what traditional air cooling can handle. Motivair’s direct liquid cooling systems enable higher compute density without compromising reliability or energy efficiency.

The Lake Mariner project illustrates a broader trend of repurposing legacy industrial sites, which often carry existing power infrastructure and large land footprints, into next-generation AI campuses. With Core42 and Fluidstack in place and the campus still expanding, TeraWulf and its partners expect capacity and demand to grow in tandem.

Schneider Electric is a global energy technology leader with around 160,000 employees and partners in over 100 countries. The company electrifies, automates and digitalizes industries ranging from buildings and factories to data centers and power grids. Its portfolio includes intelligent devices, software-defined architectures, AI-powered systems, digital services and expert advisory services.

Motivair by Schneider Electric is a leading provider of advanced liquid cooling solutions for high-density computing environments. It serves silicon manufacturers, server OEMs, colocation providers and hyperscale data centers, with a product range spanning chip-level cooling to facility-wide chiller systems.

TeraWulf builds, owns and operates digital infrastructure in the United States focused on AI and HPC. Led by veteran energy infrastructure entrepreneurs, the company delivers scalable and resilient compute capacity for major technology companies across its purpose-built campuses.

Photonic Inc has closed an investment round of more than US$200 million, giving the Vancouver-based quantum computing company a post-money valuation of US$2 billion.

The final close brings Photonic’s total capital raised to more than US$350 million, the company said on Tuesday. The round was led by Planet First Partners, a UK-based sustainable technology growth equity firm, and included new investors Business Development Bank of Canada (BDC), Export Development Canada (EDC), Bell Ventures, Firgun Ventures, InBC Investment Corp. and existing investor Mubadala Capital.

Photonic said the round’s first close, announced in January, had attracted strategic investors Royal Bank of Canada and TELUS, a Canadian telecommunications group, along with returning investors British Columbia Investment Management Corporation and Microsoft. It said the investor base reflected support across Canada, Europe, the UK, the US and the Middle East.

The funding comes as quantum computing companies race to develop machines that can move beyond laboratory demonstrations and support commercial workloads. Photonic is developing distributed quantum computing systems using what it calls its Entanglement First architecture, which combines silicon-based qubits with photonic connectivity. The company says this could allow quantum systems to scale across existing telecom infrastructure.

“This financing unites government, strategic partners, and international investors around a shared conviction: that commercial-scale quantum computing is within reach – and that its economic impact will be transformative,” said Don Mattrick, chief executive of Photonic. “We will use this funding to continue to hit key milestones, grow our team, and deepen the partnerships that will take us there,” he said.

He added that Photonic is already delivering on commitments to customers, including as part of the Canadian Quantum Champions Program and Stage B of the Defense Advanced Research Projects Agency’s (DARPA) Quantum Benchmarking Initiative.

“Distributed architectures will be an important way to scale quantum technology, and Photonic is an important partner in advancing that future,” said Zulfi Alam, corporate vice-president at Microsoft Quantum.

“Their design allows quantum systems to operate over today’s fiber infrastructure, offering a practical and scalable path toward the large-scale systems that transformative applications will demand,” he said. “We’re pleased to continue our partnership as they take this next step.”

BDC said its investment was made through its StrongNorth Fund, which backs Canadian companies developing defense and dual-use technologies. Peter Suma, managing partner of the fund, said Photonic was building scalable quantum capabilities with global partners while drawing on Canadian talent.

Bell Ventures said it would collaborate with Photonic to develop sovereign, scalable quantum computing capabilities in Canada. TELUS also pointed to the potential integration of Photonic’s architecture with its PureFibre infrastructure.

Photonic, headquartered in Vancouver with operations in the US and the UK, has more than 160 staff members. It is developing quantum computers and quantum networks for applications including materials science, drug discovery, climate change and security.

The full announcement is available here.

Ethos, a London-based artificial intelligence platform that matches skilled professionals with work opportunities, has raised US$22.75 million in Series A funding as investors back new ways to identify human expertise in an AI-driven labor market.

The round was led by Andreessen Horowitz, with participation from General Catalyst, which led Ethos’ 2024 seed round. XTX and Evantic also participated in the financing.

Ethos was founded by James Lo and Daniel Mankowitz in 2024. It uses AI to evaluate expertise through voice conversations and existing portfolios of work, including academic papers, code repositories and professional content. The company said AI has made it easier to generate CVs and job applications, but harder for organizations to identify genuine expertise.

“A CV is a poor proxy for what someone is truly capable of,” said James Lo, co-founder and chief executive of Ethos. “Ethos helps people to see the full shape of their expertise and where it fits in an AI-transformed economy - acting as their agent, opening doors they didn’t know existed, and giving companies a far more precise way to find the people they actually need.”

The platform matches users with opportunities across consulting, expert calls, market research, AI data projects and full-time roles. Ethos said more than 35,000 experts are joining the platform each week across accounting, banking, consulting, law, technology and healthcare, as well as skilled trades including electricians and plumbers.

The company said the average expert on Ethos earns £4,500 in additional income each month, while the top 10% of earners make more than £7,000 a month. Since January, the number of experts earning income through the platform has increased sixfold.

“AI is reshaping the labor market faster than our tools for valuing human expertise can keep up. Ethos is built to change that,” Lo said.

“Knowledge is our most precious resource, yet for too long the expertise inside people’s heads has been invisible to the economy,” said Jeannette zu Fürstenberg, president and managing director of General Catalyst.

“Ethos changes the game by capturing the true value of what people know and connecting it to the opportunities that need it most,” she said.

“Our Seed philosophy is simple: back exceptional leaders early,” zu Fürstenberg said. “From our first conversations with James and Daniel, it was clear they were visionaries who understood exactly where the future of work was heading and had the conviction to build for it.”

Ethos was founded by a team from Google DeepMind, McKinsey and SoftBank. Its clients include AI labs, investment funds and corporates seeking access to human expertise on demand.

monday.com has repositioned itself as an “AI Work Platform” in what the company described as the biggest change in its history, as software vendors race to embed artificial intelligence more deeply into workplace tools.

The work management software company said on May 6 that it had rebuilt its platform around collaboration between employees and AI agents, moving beyond task management into systems that can help execute work across departments.

The company said its new AI agents are built directly into monday.com and can be configured by users without technical expertise. The agents are designed to work across live business data, workflows and permissions, while remaining under human supervision.

monday.com said the agents can support tasks such as drafting marketing campaigns, qualifying sales leads, closing support tickets, onboarding new employees and processing purchase requests. The company is positioning the move as a response to a common corporate challenge: large AI investment has not always translated into production use or measurable business impact.

“Our customers are running real businesses in a world that’s changing fast, and they need a platform built for that reality,” said Roy Mann, co-founder and co-chief executive of monday.com. “monday.com is now a place where people and agents work side by side.”

The launch comes as software-as-a-service companies face pressure to prove that AI can deliver practical productivity gains rather than remain an add-on feature. The company said enterprises have widened access to AI, but many organizations still struggle to move experiments into production or use AI to transform operations.

Eran Zinman, co-founder and co-chief executive of monday.com, said the company was “going all-in” on the new strategy.

“We have 250,000 customers running their business on monday.com, and we owe them more than another AI feature,” Zinman said. “We owe them a platform built for what comes next.”

As part of the launch, monday.com is adding one-click connectors to AI platforms including Anthropic’s Claude, Microsoft 365 Copilot and OpenAI’s ChatGPT. The company said this would allow customers to bring their preferred AI systems into existing workflows.

Other updates include access to multiple large language models through monday’s AI Platform Gateway, new AI-powered development tools in monday vibe and a redesigned mobile app combining Sidekick and AI agents.

monday.com said the next phase of software-as-a-service will be shaped by companies that can turn AI into reliable business outcomes at scale.

The full announcement is available here.

Schneider Electric, an energy technology company, has signed a partnership with GreenScale to help design AI-ready data centers across the latter’s planned European campuses, as rising computing demand forces operators to rethink how facilities are built and maintained.

The companies said the collaboration will focus on reference architectures for data centers serving artificial intelligence, cloud and high-performance computing workloads. These designs will include automation, predictive analytics, condition-based maintenance and remote monitoring from the start, rather than adding them after construction.

The deal comes as AI workloads are increasing pressure on data-center operators to deliver higher power density, more reliable cooling and tighter operational control. Facilities built for conventional cloud demand are now being adapted or replaced to handle more intensive computing clusters.

“As demand for AI, Cloud and HPC accelerates in Europe, data center operators must rethink how facilities are designed and managed,” said Dan Thomas, chief executive of GreenScale.

Thomas said the partnership showed how “advanced data center architectures and digital innovation can unlock new levels of automation, efficiency and resilience.” He added that the work would help set “a new standard for intelligent design” for GreenScale’s customers.

Schneider Electric and GreenScale said the approach should reduce unnecessary maintenance, improve asset performance and lower lifecycle costs. It is also intended to give operators better visibility into equipment, allowing maintenance teams to focus on targeted interventions rather than fixed service schedules.

The companies said such systems could be particularly useful in remote or emerging regions, where fewer on-site staff, longer supply chains and higher service risks can make traditional maintenance models more difficult.

“GreenScale’s vision for its European data centers represents a new era in advanced design, where automation, efficiency, and real-time visibility are embedded from day one,” said Thierry Chamayou, vice president for cloud and service providers in Europe at Schneider Electric.

Chamayou said Schneider Electric’s role was to help create “a resilient, AI-ready infrastructure platform” that could operate efficiently in demanding environments.

The partnership will also support a unified instrumentation, monitoring and control system linking physical infrastructure with digital systems through sensors and remote tracking. GreenScale said this would help its campuses manage high-density AI clusters and cloud computing workloads across multiple locations.

The full announcement is available here.