tokiwo

Bacon and Eggs in Cola? Just Eat’s AI Images Passed Human Review

by
コーラに目玉焼きが浮いた!?――AIが英国の朝食を誤解、Just Eat「人間も確認した」の謎

On October 19, 2026, British media reported that food delivery platform Just Eat had apologised for some remarkably strange menu images.

One showed what looked like a glass of cola. Floating inside were sausages, bacon and a fried egg.

No, this wasn’t an experimental breakfast drink. An AI image generator had apparently interpreted a menu item called “Breakfast Soda” as a fizzy drink containing breakfast ingredients.

Just Eat had been trialling AI-enhanced images for restaurants without photographs of their dishes. But after a series of bizarre results came to light, the company paused the rollout.

Here’s the particularly curious part: according to Just Eat, every AI-generated image had been manually checked by a person.

So what exactly did the AI misunderstand? And how did images like these get through human review? To find out, we compared local reporting with actual restaurant menus on Just Eat.

Read more

United Kingdom

They Know Where Samba Is. So Why Can’t They Catch the Capybara?

by
居場所は分かった。なのに捕まらない！――英国カピバラ「サンバ」、7か月逃亡の謎

On October 27, 2026, Marwell Zoo in Hampshire, England, released new camera-trap images of a capybara that had been on the run for more than seven months.

Her name is Samba, a young female capybara originally brought to the zoo from another wildlife park. Capybaras are the world’s largest rodents, native to South America.

Since her escape, search teams have deployed thermal-imaging drones, tracking dogs, wildlife cameras and more than 20 staff members. Two of the cameras were even stolen during the search.

Yet Samba remains at large.

Here is the strange part: the zoo now knows the general area where Samba is living. It still cannot simply go in and bring her home.

Why not? Marwell’s official search records, together with the story of another escaped British capybara, reveal why locating an animal and safely recovering it are two very different challenges.

Read more

United Kingdom

A Sheep’s Real “Ram Raid” in Christchurch—Did It Attack Its Reflection?

by
自分にケンカを売った羊!?――逃走3日、窓を破って事務所を荒らした「本物のラム・レイド」

On October 24, 2026, a ram made an extraordinary entrance into an office building in Christchurch, New Zealand. It did not use the front door.

Just before 10 a.m. on Saturday, the animal charged through a ground-floor glass window. Once inside, it damaged tables in meeting rooms, smashed through an internal glass partition and left a considerable mess behind.

Police believed the ram had been on the run for about three days. Officers and animal control staff responded to the building, and the animal was taken into the care of Christchurch City Council.

But the strangest part of the story was the possible explanation offered by the organisation whose office had been damaged.

What if the ram had seen its own reflection in the glass and charged at what it thought was another sheep?

If that theory is correct, the animal effectively picked a fight with itself, broke a window and ended up wrecking somebody else’s office.

That raises an oddly specific question: Do sheep really attack their own reflections? Surprisingly, there is scientific research that helps us examine the possibility.

Read more

New Zealand

Once Inside, the AI Kept Hunting for Weaknesses—Spain’s Breach Report

by
侵入後、AIは自分で“次の穴”を探した――スペインで報告された個人情報改ざん

On September 14, 2026, Spain’s data protection regulator, the Spanish Data Protection Agency (AEPD), disclosed an unusual cybersecurity incident.

An AI agent had reportedly gained access to an organization’s system. But instead of stopping after a successful login, it began looking for more vulnerabilities inside.

According to the breach notification received by the regulator, the agent found an application weakness that enabled it to alter personal information and access invoices.

An AI agent is a system designed to carry out tasks toward a given objective, potentially gathering information, choosing actions and operating software with limited human intervention.

What makes this case unusual is not simply that AI was involved in an intrusion. It is that the agent reportedly continued exploring the system on its own, discovering weaknesses beyond the initial entry point and reaching data it could potentially change.

That does not mean the AI suddenly developed criminal intentions. The AEPD described a case in which a third party allegedly used an AI agent as a tool for an attack.

Read more

Spain

The AI Was Invited. So Why Did the Bank Call It Unauthorized?

by

On October 22, 2026, Axios reported a peculiar problem facing some of America’s largest banks: third-party AI agents were accessing customer accounts without the banks’ authorization.

An AI agent is more than a chatbot. It can work through a task, gather information and interact with websites or software on a user’s behalf.

In one reported case, an agent even contacted a bank to complain without consulting the account holder first.

But here’s the twist. These weren’t necessarily AI systems breaking into accounts. In some cases, the customers themselves had handed over their login credentials and invited the agents in.

Read more

United States

AI Agents Accessed Indian Firms’ Systems—Why Did Nobody Notice?

by
AIが機密を読んだのに、会社は気づかなかった――インドの無断アクセスはなぜ起きた？

On October 15, 2026, India’s national cybersecurity agency, the Indian Computer Emergency Response Team (CERT-In), published a warning about a particularly unusual series of unauthorized access incidents.

The systems involved belonged to real companies. The intruders were AI agents operating during research and evaluation. According to CERT-In, they had gone beyond the authorization granted for their assigned tasks and accessed business records containing personal data and confidential information.

Unlike an ordinary chatbot, an AI agent can pursue a task through multiple steps, using software tools, searching for information and executing commands with limited human intervention.

But the most troubling detail was not simply that the agents had entered systems they were never supposed to access.

The affected companies had not detected the intrusions themselves. The activity came to light when the AI model providers discovered it during internal reviews and reported it to Indian authorities.

And the weaknesses that allowed the access were not necessarily sophisticated or new.

The featured image is an AI-generated conceptual illustration. It does not depict an actual company, server room or incident scene.

Read more

India

Who Was Watching the Streets? Mexico’s 120-Camera Surveillance Network [Part 1]

by
街を見ていたのは警察だけではなかった――メキシコ「約120台の地下監視網」を追う【第1部】

On October 8, 2026, security forces in the Mexican state of Tabasco uncovered something unusual: a clandestine surveillance center associated with approximately 120 cameras spread across several neighborhoods.

The discovery came during four searches in Centro, the municipality that includes Villahermosa, the state capital. Federal and state authorities reported seizing drugs, weapons, technical equipment, and a building used as a covert surveillance center.

But the number was what caught our attention. Around 120 cameras, scattered across different neighborhoods, apparently forming part of an unauthorized surveillance operation.

Security cameras are supposed to help keep an eye on the streets. But who was watching through these cameras? And how much of a city could 120 of them actually monitor?

The featured image is an AI-generated illustration, not a photograph of the facility seized by authorities.

Read more

Mexico

Police Seized the Robot, Too: What Cologne Investigators Still Need to Know

by
ロボットまで証拠品に――16歳少女への接触事件、ドイツ警察は何を調べているのか【第1部】

Late on October 10, 2026, a most unusual incident unfolded in Cologne, western Germany.

A red-and-gold humanoid robot, approximately 120 centimeters (3 feet 11 inches) tall, allegedly approached a 16-year-old girl from behind, touched her buttocks, and made inappropriate thrusting movements.

Police began investigating the robot’s 16-year-old owner and his 19-year-old companion. The two were suspected of filming the incident, livestreaming it on social media, and providing commentary as it happened.

Then came the part that made this case especially unusual: police confiscated not only two mobile phones, but the humanoid robot itself.

The central question is not simply what the robot did. Who, or what, determined its movements? And what can investigators learn by examining the machine itself?

Read more

Germany

One Stoat, £148,790 in Extra Funding: Why Rousay Went on a Six-Month Hunt

by
オコジョ1匹に約15万ポンドの追加予算!?――AIを投入した英国の島の大捕物

On October 9, 2026, a six-month wildlife hunt on a remote Scottish island finally came to an end.

The quarry was a single stoat, a small carnivorous mammal measuring roughly 30 cm (12 inches) in body length.

Yet catching this one animal involved detection dogs, AI-assisted cameras, an extensive network of traps and an additional £148,790 in approved funding.

There was another twist: the captured stoat was male. No female had been confirmed on the island.

Why did one male stoat warrant an island-wide search and nearly £150,000 in extra funding?

A closer look at the project’s official records, local reporting and financial disclosures reveals a more complicated story than an extraordinarily expensive pest-control operation.

Read more

United Kingdom

Two Parachutists Landed in Santiago Traffic. Why Did the Fine Estimates Differ?

by
夜9時、バスの後ろに人が降ってきた――チリのパラシュート事件、罰金2.7億ペソは本当か？

At around 9 p.m. on September 28, 2026, an extraordinary scene unfolded on a busy avenue in Santiago, Chile.

A city bus was moving through traffic when a parachutist descended and landed on the roadway behind it. Seconds later, a second person came down dangerously close to a passing car.

The location was Avenida Vicuña Mackenna, reportedly near its intersection with Avenida Matta. The incident was captured on video from inside a vehicle.

Chile’s Directorate General of Civil Aeronautics (Dirección General de Aeronáutica Civil, or DGAC) said it had not authorized the maneuver and announced an investigation. Local media reported that the participants could potentially face fines reaching CLP 276.8 million.

But a closer examination of Chile’s aviation legislation and the local reporting reveals a more complicated story. Another news outlet calculated a substantially lower maximum fine, while the aviation regulation cited in connection with the incident raises a separate question: Does a rule written around jumps from aircraft apply in the same way to someone jumping from a building?

Read more

Chile