FlashFeed

Full text

Enterprise companies are running AI agents ahead of the controls needed to manage them — and they deployed that way knowingly. That is the central finding from VentureBeat Research's June survey of 573 technical leaders at companies with 100 or more employees, fielded across five parallel surveys of the agentic stack.

Enterprises are now retrofitting to catch up with their own standards, and they are budgeting for it: Roughly six in 10 enterprises plan to switch or add vendors in each of five control layers within the next 12 months, and roughly a third — depending on the layer — plan to move within the quarter, the research finds.

There are five main layers where enterprises are building: identity for agents (which agent is allowed to do what, under whose credentials); evaluation of agent output (whether the work is any good); cost telemetry (what each agent costs to run); the context layer (the business data and definitions agents draw on to answer); and the orchestration control plane (the software that coordinates multi-step agent work).

Enterprises are already paying the price for deploying agents ahead of adequate control functions. Fifty-four percent of companies had an agent security incident or near-miss caught before harm in the past 12 months. Twenty-seven percent exercise only reactive control of agent spend — they learn what an agent costs when the invoice arrives, with no per-agent budget or ceiling in place.

Here are the five findings that anchor the set — one finding per layer of the tech stack — and what the data suggests doing first in each.

Expensive hardware is idle: 86% of GPU operators report utilization of 50% or less

Eighty-six percent of enterprises that run their own GPUs report utilization of 50% or less. Wall Street has spent the quarter debating whether the AI buildout is overbuilt. This is buy-side measurement, from the enterprises doing the buying, and the research says the most expensive hardware in buildings of these enterprises runs at no more than half its capacity.

The measurement gap compounds it: A minority 44% rigorously track what their AI compute actually costs and returns. Everyone else is only estimating. And the enterprise shopping process continues regardless: 45% of these enterprises say the emerging compute option they are most likely to evaluate in the next 12 months is an AI-specialized cloud (CoreWeave, Lambda, Crusoe, Nebius). However, under 2% of these enterprises report using one of these neoclouds today.

Moreover, roughly one in three companies appears to be considering a hedge against Nvidia: Asked which emerging compute option they are most likely to evaluate in the next 12 months, 32% of enterprises named non-Nvidia accelerators (AWS Trainium, Google TPUs, AMD), while 28% named next-generation Nvidia GPUs. The data suggests that enterprises should measure the utilization and per-workload cost of the GPUs they already own before committing budget to new compute — whether that's an AI-specialized cloud contract, new accelerators, or more GPUs.

Most deployed "agents" do single-prompt work: 71% say a quarter or fewer complete multi-step tasks on their own

Seventy-one percent of enterprises say a quarter or fewer of their deployed "agents" can complete multi-step work on their own; the rest are single-prompt chatbots. Only 10% say true agents are the majority of what they run. To be sure, the respondents reported that they are in a position to know these things: 81% said they recommend or decide AI purchases at their companies.

That finding — that most agents are actually just chatbots in trenchcoats — lands amid adoption claims across the industry running well ahead of what enterprises are actually running. Gartner predicted 40% of enterprise applications will be integrated with task-specific AI agents by the end of 2026, up from less than 5% in 2025. It also warned that the most common misconception is referring to these AI assistants as agents, a misunderstanding known as "agentwashing."

Meanwhile, Zapier's enterprise survey said 72% reported deploying or testing autonomous agents; and Writer's 2026 survey has 97% of executives saying their company deployed AI agents in the past year.

Those surveys asked whether companies have deployed something called an AI agent, and companies said yes. Our survey asked the people running those deployments a harder question: Of the agents you have in production, how many can complete a multi-step task without a person driving each step? The gap matters for two practical reasons. First, the inflated adoption figures are the benchmark boards and vendors use to pressure technical leaders into moving faster — and this data says the real bar is far lower than the headlines suggest. Second, the label determines the bill: A single-prompt chatbot with a human reading every answer needs none of the identity, evaluation, and cost controls this report covers, while a true multi-step agent needs all of them.

66% let agents push to production on automated evals alone — or are engineering toward it. 5% fully trust those evals

Two-thirds of enterprises fall into one of two camps: 34% already allow an AI agent to push a code or system change to production based on automated evaluation results alone, with no human reviewing it, and another 33% are actively engineering their pipelines to allow that within the next 12 months. Only five percent fully trust the automated evaluations that would make that decision.

The distrust is earned. Half of enterprises shipped an agent that passed internal evaluations and then caused a customer-facing failure in the past year; a quarter watched it happen more than once. Asked to name the biggest weakness in their current evaluations, more enterprises chose “poor alignment with real-world outcomes” than any other answer — 29% of respondents.

And most of the checking happens before an agent ships, then stops. Once agents are live with real users, only 23% of enterprises run real-time quality checks on the answers those agents produce. Another 51% monitor system health only — uptime, request traces, and gateway logs — which tells them the agent is running, and nothing about whether its answers are right. The first move: Before removing human review from any workflow, test your evaluations against production outcomes rather than internal benchmarks, and instrument answer quality, not just uptime.

This finding is explored in more depth in VentureBeat's related coverage of the evaluation gap , which found that larger enterprises are moving faster toward zero-human deployment while also failing more often — and outlines a regression-testing framework built on production outcomes rather than internal benchmarks.

69% run credential sharing somewhere in the agent fleet — and those companies get hit far more often

Sixty-nine percent of companies allow agent credential sharing somewhere in their agent fleet during runtime – meaning multiple agents operating under one API key or service account. Those companies were far more likely to get hit: Organizations with credential sharing anywhere in the fleet experienced a security incident or near-miss at a 63.5% rate (47 of 74), against 40.9% (9 of 22) where every agent has its own scoped identity.

The takeaway for enterprises is this: Give every agent its own scoped identity, starting with the agents that touch production systems.

57% traced a confident, wrong agent answer to their own missing or inconsistent business context

Fifty-seven percent of enterprises traced at least one confident, wrong agent answer in the past six months to missing or inconsistent business context: wrong metrics, stale definitions, absent documents. Most of them watched it happen more than once.

Most enterprise companies are fixing this, even though they’ve moved forward with agent deployment already: 25% already run a governed semantic layer, or one governed definition of the business that every AI reads from, in production. However, 34% are still building one, and 41% haven't started. The takeaway: Govern the definitions your agents answer from, metrics and entities first, before scaling the agents that depend on them.

The quarter where agent technology “portability” became a priority

One more shift is worth reporting with its limits stated plainly. In our spring orchestration survey wave, the top concern about provider-controlled orchestration was security and permissioning limits (32%). By June, vendor lock-in led at roughly a third, with security limits at 28%.

Those are two snapshots one quarter apart, and here’s one possible explanation for why portability became a top issue for enterprises. Our June survey went into market after a June 12 U.S. Commerce Department export order took Anthropic's Claude Fable 5 offline for enterprises for roughly three weeks. Meanwhile, Chinese company Z.ai released GLM-5.2's open weights under an MIT license on June 16 at roughly one-sixth of GPT-5.5's price; and Tencent's Hy3 arrived July 6 under Apache 2.0; and OpenAI previewed GPT-5.6 on June 26 to a small group of government-vetted partners, opening it broadly on July 9 after the government's review cleared. The open-weight releases in particular promise enterprises more control over their agents, and while we haven't established a causal link here, the timing is worth noting.

The posture data matches the mood: 51% now expect their primary control plane for enterprise agents to be hybrid — provider-native plus external orchestration — by the end of 2026, up from 34% in the spring survey wave. Enterprises reporting that they rely purely on provider-managed agent services fell from 12% to 7%.

Five layers, no incumbents, 12 months

The synthesis across all five surveys reveals a huge “buying” window. In each of the five control layers, 57% to 64% of enterprises plan to switch or add vendors within 12 months — 64% in infrastructure and in evaluations, 59% in agent security, 57% in retrieval and context — and 26% to 38%, depending on the layer, plan to move within a quarter. No layer has an established incumbent: The most common evaluation tooling is the model provider's built-in evals, tied with no dedicated tooling at all (17% each); 82% of respondents name provider-native or hyperscaler controls as their primary agent security layer; and provider-native retrieval leads the context technology layer (RAG, etc) as well.

Most enterprises are defaulting today to the built-in tools that ship with the big AI platforms they already use: Anthropic, OpenAI, Google, Microsoft, and AWS. That holds true across every one of these agentic technology layers: enterprises are looking to their primary cloud and model providers to supply the guardrails, evaluations, and retrieval solutions already bundled into those providers' offerings.

Those defaults are winning on convenience, and they're also what the coming spending decisions will test. The survey didn't ask which direction that money moves — toward the platforms' built-in tools or toward the specialists challenging them — which is exactly why every contract in these five layers is worth watching over the next four quarters.

The Q3 survey wave will measure whether the enterprises made good on these budget plans: whether their agents gained scoped identities, whether evaluations got tested against production outcomes, whether GPU utilization rose, and whether the semantic layers under construction shipped.

VentureBeat will release the full Q2 reports across all five VB Pulse trackers at VB Transform , July 14–15 at Hotel Nia in Menlo Park, where we convene enterprise technical leaders building autonomous agents in production.

Disclosure: VentureBeat produces both this research and VB Transform

Full text

Early Christian communities in Sweden often buried children in the same grave with adults, but archaeologists have found that these individuals rarely shared close biological ties, raising the question of how medieval people interred their dead.

In a new study, researchers analyzed the DNA of 142 skeletons from three cemeteries in Sweden dating to the 10th to 14th centuries, focusing on collective burials in which two or more people were buried in the same tomb.

"We often assume that adults and children sharing a grave were parents and children or other close family members," study first author Maja Krzewińska , a paleogeneticist at Stockholm University, said in a statement . "In most cases, that was not what we found."

The researchers determined that most burials containing multiple individuals held both adults and children and that the people buried together were usually of the same sex — a woman buried with a girl or a man buried with a boy. But the DNA analysis held a surprise: People buried together rarely exhibited close biological kinship, the researchers wrote.

When Christianity spread across Scandinavia starting in the late 10th century, burial practices became more uniform. Graves were oriented east to west, and people were buried in a simple shroud without any grave goods. Baptized individuals were allowed to be interred in consecrated cemetery grounds, while babies who died before they could be baptized were excluded.

"We have previously analyzed a burial containing an adult and the remains of a fetus, which we believe represents an unbaptized individual," Krzewińska told Live Science in an email.

A photograph of the Västerhus church ruin, Frösö parish, in Jämtland, Sweden, before 1951, where archaeologists have found many burials of children who weren't interred with close family members. (Image credit: Riksantikvarieämbetets arkiv ) These unusual burial arrangements likely point to early Christian traditions. For instance, some of the children buried with adults in the new study may have been unbaptized. Normally ineligible for burial in the cemetery, the children may have been opportunistically interred with an adult to get around religious norms. Other burials may reflect unrelated people buried together in the spring after passing away in the winter, when burial in the frozen ground was impossible.

"We also believe, based on more distant genetic affinity, that some co-burials represent more distant family relations, or even non-biological kin group relations," Krzewińska said.

In early medieval Scandinavia, households often included extended relatives, servants, employees and enslaved individuals, the researchers wrote. While biological kinship played a large role in the organization of society, membership in the local Christian community may have been equally important in determining where and with whom to bury a deceased person.

"Archaeologists have debated the relationships between people buried together in this type of grave for a long time," study co-author Anna Kjellström , an archaeologist at Stockholm University, said in the statement. "Ancient DNA has finally given us the tool we have been waiting for to test these interpretations directly."

We are family

In addition to collective burials of unrelated people, the archaeologists discovered evidence that some families were buried within the same cemetery over several generations. One burial, known as Lady 56, was a Christian pilgrim who anchored three generations of relatives.

A photograph of a pilgrim shell found at the Västerhus cemetery. This type of scallop shell is a symbol of Christian pilgrimage to Santiago de Compostela in northwestern Spain. (Image credit: Christer Åhlin, Historiska museet, 2012 ( CC BY 4.0 )) Lady 56 died when she was around 30 years old. She was buried with a rare scallop shell, a symbol of the apostle James, that she obtained after completing a pilgrimage to Santiago de Compostela , a town in northwest Spain on the edge of Christian Europe.

The researchers also identified Lady 56's kin group, which was of particular importance to the community and stretched over generations in the Västerhus cemetery, Krzewińska said. The DNA analysis revealed that Lady 56's parents, brother and daughters were also buried in the same cemetery, but in different places.

Related Stories

1,000-year-old burials of 'first Christians' in Poland discovered near medieval settlement

Medieval grave of 'very, very powerful' man and his 4-foot-long sword unearthed in Sweden

Up to 20,000 coins from Early Middle Ages discovered by man digging for worms near Stockholm

Västerhus was part of a wealthy landowner's farm from the 11th to 14th centuries, and the cemetery contained the remains of more than a dozen members of a biologically related group, many of whom were interred with members of a different kin group.

The DNA connections between the main Västerhus family and other kin groups in the cemetery support the special status of the main family. These close-kin burials highlight the importance of ancient-DNA testing, as different burials from the same time and region can follow very different traditions.

What do you know about Jesus Christ, the man? Test your knowledge of biblical archaeology with our Jesus Christ quiz!

Full text

Keep our news free from ads and paywalls by making a donation to support our work!

Notes from Poland is run by a small editorial team and is published by an independent, non-profit foundation that is funded through donations from our readers. We cannot do what we do without your support.

Soldiers in Poland will be able to pick up their uniforms from their nearest parcel locker rather than having to collect them from military stores under a new scheme being rolled out by the armed forces, with private logistics giant InPost responsible for deliveries.

The move is a “simple, practical change” that will make life easier for troops, said a deputy defence minister. The CEO of InPost, which is Europe’s largest operator of parcel lockers, praised the decision to make “dual use” of private infrastructure for national-security purposes.

📦 Żołnierze odbiorą umundurowanie w paczkomacie. Trwa pilotaż nowego systemu.‼️

🔊W Siłach Zbrojnych RP trwa pilotaż nowego systemu, który rozszerza możliwości zaopatrywania żołnierzy. Już od przyszłego roku funkcjonować będzie Platforma Wysyłkowa Przedmiotów Umundurowania i… pic.twitter.com/NGUAQognYX

— Inspektorat Wsparcia Sił Zbrojnych (@IWsp_SZ) July 9, 2026

Under the new system, military personnel can order equipment via an online system and collect it from parcel lockers, which are devices that allow users to collect packages independently at any time.

“We want soldiers to be able to order their equipment in a modern, simple, and convenient way, without having to adjust to the opening hours of stores, and to be able to do so at any time they choose,” said General Dariusz Mendrala, the head of the Armed Forces Support Inspectorate (IWsp SZ).

A small pilot scheme began last year and has now been expanded to over 20,000 soldiers. That figure will rise to around 60,000 by the end of this year, before the system is fully rolled out in 2027.

Click here to help us continue providing news free from paywalls and ads

“This is a simple, practical change…that will make soldiers’ daily service easier,” said Cezary Tomczyk, a deputy defence minister, sharing details of the scheme on social media.

He indicated that the system covers other equipment in addition to uniforms, though neither he nor IWsp SZ have specified what types of gear are included.

Sharing Tomczyk’s post, Rafał Brzoska, the founder and CEO of Inpost, said that the scheme was a good example of the “dual use” of civilian and commercial technologies for national security purposes.

“Polish business has been calling for years: let’s use domestic technologies, infrastructure and enterprise to reinforce the state’s defence,” wrote Brzoska. He noted that Ukraine has shown, for example through its drone industry, how private enterprises can support the military.

Polski biznes od lat apeluje: korzystajmy z krajowych technologii, infrastruktury i przedsiębiorczości do wzmacniania obronności państwa. @MON_GOV_PL właśnie pokazuje, jak przejść od deklaracji do działania!!

I to jest właśnie realny dual use, o którym wszyscy mówią, tylko w… https://t.co/oXA8dmoVpE

— Rafał Brzoska (@RBrzoska) July 10, 2026

A spokesman for IWsp SZ, Lieutenant Colonel Piotr Płuciennik, confirmed to broadcaster TVN that InPost had been selected in a tender to manage the deliveries until December this year. He said that the system was set to cost around 2 million zloty (€462,00) a year.

InPost operates by far Poland’s largest network of parcel lockers: its almost 29,000 devices represent around half the total number in the country. The company has also expanded rapidly in western Europe, where it operates parcel lockers in the UK , France and Spain , among other countries.

Płuciennik said that the delivery company is given pre-packaged parcels that feature only the soldier’s phone number on the label. For security reasons, they are delivered only to parcel lockers, and not to home addresses.

With an estimated 233,800 military personnel in 2025, Poland’s armed forces are the third-largest in NATO , behind only the United States and Turkey. That figure has more than doubled in the last decade as Poland has ramped up defence spending in response to a growing threat from Russia.

Poland has NATO’s third-largest military and the alliance’s largest in Europe, new figures show.

For more, see our full report here: https://t.co/QBBVQ9cqzS pic.twitter.com/TA34KQrJ4k

— Notes from Poland 🇵🇱 (@notesfrompoland) July 16, 2024

Notes from Poland is run by a small editorial team and published by an independent, non-profit foundation that is funded through donations from our readers. We cannot do what we do without your support.

Main image credit: st.szer. Wojciech Król/CO MON (under CC BY-SA 4.0 )