Astronomers just received the largest and sharpest visible-light portrait ever assembled of the Milky Way’s central bulge. The European Space Agency’s Euclid telescope produced this six-gigapixel mosaic during a single day of observations in March 2025, packing more than sixty million stars into one frame along with dark dust clouds and pockets where new stars are forming.

The image spans a vast expanse of sky that most space telescopes cannot capture in a single glance. To get the complete image, 9 separate snaps from Euclid’s camera were stitched together, with each section covering more ground than the entire Moon from Earth. The original data was black and white, but colors were added later using similar observations from the Canada France Hawaii Telescope, which greatly improved the identification of different types of stars and gas.

LEGO Art The Milky Way Galaxy Building Set – Space Decor DIY Kits for Adults for Bedroom wall, Ages…

• Milky Way wall art décor – Create a large Milky Way Galaxy piece of art with this LEGO building kit for adults
• LEGO 3D-effect wall décor – A Milky Way Galaxy artwork created from 3,091 layered LEGO bricks and pieces to produce a 3D effect with depth and…
• Packed with features – Wall art includes some of the Milky Way’s most famous stars, clusters and other spectacles, including Trappist-1, The…

The middle of the frame is covered in golden yellow stars that are so tightly packed that it resembles a sparkling sprinkle of glitter on sand. The galactic bulge, a massive center structure containing 8 billion stars, is primarily composed of older, colder stars. Bits and pieces of deeper colors and channels cut through everything like black ink blots or wisps of smoke. These are dense clouds of dust and gas soaking up the light from the stars behind them.

As you move up the image, the color palette changes slightly, with the reds and purples becoming more stronger and some dazzling blue lights standing out against a faint red glow. These blue lights are actually young, enormous stars that recently formed in one of the Milky Way’s spiral arms. Their light ionises the hydrogen gas around them, resulting in the red glow seen in spots. Everything is visible from a distance of approximately 26 000 light years away, and you’d have to go through a lot of intervening material in our galaxy’s disk to see it.

The amount of detail here is incredible, especially given how bright and packed the galactic center is generally, which swamps the detectors with all that brightness and dust, but Euclid’s built-in sharpness allows it to separate individual stars even in the most congested locations. The resolution is equivalent to the Hubble camera, but it can capture 270 times more sky in each frame.

A ground-based telescope like Keck would take a whopping 2,000 hours to cover the same ground with the same level of quality. To be honest, the true value of this shot stems from using it as a baseline for future observations. Astronomers will be able to compare it to later images to discover microlensing events, which occur when the gravity of a foreground star and any planets it may have temporarily increases the light from a background star.

The way this brightens allows you to tell if a planet exists, and repeating the process allows you to compute the planet’s mass. Over the last 20 years, astronomers have discovered over 300 exoplanets near the galactic core. Euclid’s map of the stars presently contains 51 known planets, and it will aid in the finding of many more, as well as determining the masses of planets that have already been detected, such as one icy globe that has been present for the past 20 years.

Ford executives said they have hired 350 veteran engineers — some of them were former employees, while others had been working at suppliers — after artificial intelligence and automated systems failed to deliver the desired quality level.

Bloomberg reports the company’s chief operating officer Kumar Galhotra told journalists that Ford had been “relying more and more on automated quality systems” with disappointing results. So the company “brought back technical specialists,” and those specialists “hunt for failure points before a part ever reaches the plant floor.”

Charles Poon, Ford’s vice president of vehicle hardware engineering, added, “Mistakenly we thought that by just introducing artificial intelligence and ingesting the design requirements that we had, that that would produce a high-quality product.”

To be clear, this doesn’t mean Ford is abandoning its AI plans entirely. Instead, it’s using the rehired employees — referred to as “gray beard” engineers — to train younger staff and reprogram AI tools.

This rehiring seems to be paying off, with Ford anticipating that it will lead to $1 billion in reduced costs this year. The automaker also claimed the top spot among mainstream brands in the JD Power Initial Quality Survey released this week.

ISTE+ASCD — the organization behind the editorially independent news site EdSurge — announced a new official name on June 28: The International Society for Transforming Education.

The announcement was made at the opening general session of the organization’s annual conference in Orlando, Florida. Jeremy Owoh, president of the International Society for Transforming Education and superintendent of Jacksonville North Pulaski School District, explained that the name change had been in the works for more than a year.

“We knew that the [merger] needed to happen first and then once we grew together as a community then we could take on that [renaming] task,” he said. “This is a change we’re making very thoughtfully.”

Richard Culatta, CEO of the International Society for Transforming Education, told attendees that the name change is intended to reflect a global focus on aligning instructional strategy, technology use and educator practice to improve student outcomes and engagement. 

“We believe that this name most effectively captures what both legacy organizations were always about,” he said. “Our new name shifts the focus from how we do it, to why we do it. And it shows how serious we are about transforming learning together.”

Some attendees expressed enthusiasm over the new name. “Oh, I’m excited,” said Elizabeth Diamond, an associate professor at Temple University in Philadelphia. “Words are so important, and those words are where we’re headed as teachers.”

Julie Keller, also a Temple University associate professor, added, “There’s power in the words, and it really brings together what we’re trying to do.”

Other attendees, including legacy-ASCD member Ruth Letang-Horton, vice president of the North American Division of SDA, were less enthusiastic. “I feel like the ASCD part is really lost,” she said. “Your feeling is like, ‘Wait a minute, what about ASCD?’ It’s because I’ve been an ASCD member for decades.”

The new name is the latest phase of the merger between ISTE and ASCD, which happened in 2023. According to Culatta, membership, educator certifications, the ISTE Standards and professional learning programs will continue without interruption under the new brand.

Read the full press release here.

(Editor’s note: EdSurge is an editorially independent newsroom of the International Society for Transforming Education.)

science

Remove something from the internet? You can’t stop the (climate change) signal, Mal

It’s back! After Donald Trump shuttered the National Oceanic and Atmospheric Administration’s (NOAA) Climate.gov website in 2025, cutting off public access to its 15-year archive of climate information, former members of the site’s team have brought much of it back at a new domain.

“Trusted climate information should not disappear when politics change,” Climate.us managing director Rebecca Lindsey said of the new platform in a press release.

Lindsey, who previously served as the Climate.gov program manager and lead editor, told The Register in an email that she and one of the web developers responsible for the site were the first to be caught up in government purges when DOGE swept through the department in late February 2025. 

“In May, political appointees directed that all the remaining Climate.gov editorial and GIS/data visualization staff be removed from the contract,” Lindsey added. 

Created in cooperation with sustainability nonprofit accelerator Multiplier, Climate.us aims to be an independent alternative to its old .gov, and many of the former NOAA crew behind the previous website have teamed up for the new initiative to “keep climate information accurate, accessible, scientifically rigorous, and useful for the people who rely on it.” 

Climate.gov, which now redirects to a NOAA page about climate but which hosts none of the data the shuttered site used to contain, was taken offline in July 2025 following a Trump executive order prioritizing “gold standard science.” The order decried what it called the prior administration’s politicization of science by, among other things, “encouraging agencies to incorporate diversity, equity, and inclusion considerations into all aspects of science planning, execution, and communication.” 

The EO called out climate change science as a particular area of concern, arguing that prior climate science models relied on worst-case scenarios, which somehow meant the public availability of 15 years of climate data and reporting ought to change. 

The shuttering of climate.gov followed a day after the order, leading to scientists expressing concern about the ability of governments, the public, and private organizations to combat the effects of a changing climate, whether the Trump administration believed the data was true or not. 

“This is evidence of serious tampering with the facts and with people’s access to information, and it actually may increase the risk of people being harmed by climate-related impacts,” University of Arizona climate scientist Kathy Jacobs told The Guardian in July 2025, following closure of the site and removal of other climate information from public repositories. 

Changes to the site actually began before that, Lindsey told us. Prior to her termination, the Climate.gov team was ordered to search its archives and remove any information that violated Trump’s Gulf of America order and ban on DEI programs. Guides on teaching climate change and principles of climate literacy were among documents purged from the site in that sweep. 

Climate.us, and Climate.gov before it, are designed to be a bridge between scientists studying the climate and the public, Lindsey told us.

“Most of those functions we can perform almost as well outside of the federal domain as in it,” Lindsey said. “However, losing access to the tremendous store of knowledge and expertise possessed by federal scientists, with whom we partnered to make sure our content was accurate, is a real blow.” 

All of the content that was purged from the .gov is now back, along with blogs from experts, climate status reports, maps and data pathways, and national assessments of climate change as well. 

Lindsey told us that rapidly changing political winds have led her to believe that the government isn’t the right place for that mission to continue, and that she would have concerns about returning the site to federal management if a future administration changed its position on climate change. 

“I believe that fostering climate literacy is a public good, one of those things that benefit society as a whole, rather than one company or person,” the Climate.us director told us. “So I would definitely have concerns that going back to the government would just put us on a hamster wheel, where we’d face the same situation the next politics shift.” 

Regardless of whether that offer comes, Lindsey said that the Climate.us team will continue with the same mission it had before the Trump administration attempted to quash it: Getting climate science in front of the public in a manner that’s understandable so they can make their own decisions about how to respond. 

“We aren’t trying to tell people what to do about climate change,” Lindsey said. “We just think that people will come up with better strategies to confronting the world’s climate challenges if they understand what the science is telling us.” ® 

The fine can now potentially hit 99 million AUD, or $68 million.

In the past two years, businesses have been trying to fit large language models (LLMs) into support, analytics, development, and internal automation like never before.

Along with the increasing adoption of AI technology, another trend is gaining momentum — cybercriminals are taking advantage of the disconnect between assumptions about LLMs and their actual characteristics.

In 2025 and 2026, several independent sources have highlighted the same trend: Prompt injection remains one of the most impactful and widely demonstrated attack vectors against LLM systems. The OWASP LLM Top 10 (2025) lists prompt injection as LLM01, identifying it as the most critical category of LLM‑specific vulnerabilities, for the second consecutive edition. OWASP’s ranking reflects the fact that LLMs still struggle to reliably separate instructions from data, making them susceptible to manipulation through crafted inputs.

CrowdStrike’s 2026 Global Threat Report — built on frontline intelligence across more than 280 tracked adversaries — documented that threat actors injected malicious prompts into legitimate generative AI tools at more than 90 organizations in 2025. They then used those injections to generate commands that stole credentials and cryptocurrency. The report stated it plainly: “Prompts are the new malware.” AI-enabled adversaries increased their overall attack volume by 89% year-over-year, with prompt injection working as both an entry point and a force multiplier.

Real‑world incidents illustrate the operational impact. In August 2024, researchers at PromptArmor disclosed a prompt injection vulnerability in Slack AI that allowed an attacker to exfiltrate data from private Slack channels they had no access to — including API keys shared in private developer channels — by placing a malicious instruction in a public channel or embedding it in an uploaded document.

In June 2025, researchers at Aim Security disclosed EchoLeak (CVE-2025-32711, CVSS 9.3), the first documented zero-click prompt injection exploit against a production AI system, targeting Microsoft 365 Copilot. By sending a single crafted email, no user interaction required, an attacker could cause Copilot to access internal files and transmit their contents to an attacker-controlled server.

Both vulnerabilities were patched. These incidents underscore the fact that prompt injection is not a theoretical weakness but a practical, repeatable threat organizations must address as they deploy AI systems at scale.

Prompt injection techniques have undergone major evolutions over recent years, now targeting multi-agent architecture, retrieval-augmented generation (RAG) pipelines, model routers, and long-term memory capabilities.

The enterprise challenge: Too much trust

Businesses deploy LLMs to process instructions, summarize information, and trigger automated workflows, but it is difficult for LLMs to tell:

This creates an opportunity for attackers to manipulate and influence the model’s behavior, either directly or indirectly.

Modern prompt injection

Cross-model prompt injection

LLM use is a common practice among enterprises. Attackers corrupt the output of a particular model, knowing well that other models would be processing the content. Hence, the corruption propagates through all AI systems.

RAG supply chain poisoning

Attackers create malicious information — documentation, blog articles, GitHub READMEs. Then they wait until this malicious information is ingested in enterprises’ RAG pipelines, then use it as an attack vector.

Agent hijacking

AI agents have evolved to the point where they can send emails, modify cloud infrastructure, execute code snippets, and interact with internal corporate systems. It takes just a single instruction to make agents act differently in a harmful manner.

Context overflow attacks

With the help of million-token context windows, attackers place malicious code within the document and hope that an LLM will stumble upon it and execute it, thus overriding all previous instructions.

Memory poisoning

Due to the implementation of long-term memory in LLMs, attackers can inject instructions that permanently reconfigure their state.

Model‑router manipulation

Enterprises increasingly use model routers to select between multiple LLMs. Attackers craft prompts that force routing to the weakest or least‑guarded model.

Why this matters for business leaders

Prompt injection is not a theoretical problem. It directly affects:

• Customer‑facing systems (chatbots, support agents)

• Internal copilots (developer tools, security assistants)

• Automation workflows (ticketing, cloud operations, HR processes)

• Data governance (RAG pipelines, knowledge bases)

The risk is no longer limited to “the model said something it shouldn’t.”

In 2026, prompt injection can:

• Trigger unauthorized actions

• Leak sensitive data

• Corrupt internal workflows

• Manipulate analytics

• Alter business logic

• Compromise multi‑agent systems

The attack surface has expanded dramatically.

What enterprises should do now

1. Constrain model permissions

Limit what the model can do, not just what it should do.

2. Segment untrusted content

Treat all external data — including RAG sources — as potentially hostile.

3. Monitor tool invocation

Require human approval for high‑impact actions.

4. Validate content provenance

Ensure RAG pipelines don’t ingest poisoned external content.

5. Harden model routers

Prevent attackers from forcing routing to weaker models.

6. Treat LLMs as untrusted components

This mindset shift is the foundation of modern AI security.

The bottom line

Prompt injection remains the most effective way to compromise enterprise AI systems because it exploits the fundamental way LLMs interpret text. Until organizations treat LLMs as untrusted interpreters — not autonomous decision‑makers — prompt injection will continue to dominate the AI threat landscape.

Julie Brunias is an AI Security Architect.

Welcome to the VentureBeat community!

Our guest posting program is where technical experts share insights and provide neutral, non-vested deep dives on AI, data infrastructure, cybersecurity and other cutting-edge technologies shaping the future of enterprise.

Read more from our guest post program — and check out our guidelines if you’re interested in contributing an article of your own!

Looking for the most recent Wordle answer? Click here for today’s Wordle hints, as well as our daily answers and hints for The New York Times Mini Crossword, Connections, Connections: Sports Edition and Strands puzzles.

Today’s Wordle puzzle came together fairly quickly for me. If you need a new starter word, check out our list of which letters show up the most in English words. If you need hints and the answer, read on.

Read more: New Study Reveals Wordle’s Top 10 Toughest Words of 2025

Today’s Wordle hints

Before we show you today’s Wordle answer, we’ll give you some hints. If you don’t want a spoiler, look away now.

Wordle hint No. 1: Repeats

Today’s Wordle answer has no repeated letters.

Wordle hint No. 2: Vowels

Today’s Wordle answer has two vowels.

Wordle hint No. 3: First letter

Today’s Wordle answer begins with C.

Wordle hint No. 4: Last letter

Today’s Wordle answer ends with E.

Wordle hint No. 5: Meaning

Today’s Wordle answer can refer to something constructed in a basic or rough way.

TODAY’S WORDLE ANSWER

Today’s Wordle answer is CRUDE.

Yesterday’s Wordle answer

Yesterday’s Wordle answer, June 28, No. 1835, was EMCEE.

Recent Wordle answers

June 24, No. 1831: QUEER

June 25, No. 1832: UNITY

June 26, No. 1833: ACUTE

June 27, No. 1834: SCOOP

Streaming ads might be getting a lot quieter this week.

A California law banning streaming services from showing ads “louder than the video content” that they accompany is set to take effect on Wednesday, July 1. (Existing legislation already imposes similar volume restrictions on broadcast and cable TV commercials.) 

Ars Technica notes that streaming services have not shared additional details about how they plan to comply with the law. While the volume limitations only apply to California for now, it seems likely that any relevant changes would be deployed more broadly, especially with a similar bill set to take effect in Illinois next year.

When the law was passed in 2025, its sponsor, State Senator Thomas Umberg, said it was inspired by “every exhausted parent who’s finally gotten a baby to sleep, only to have a blaring streaming ad undo all that hard work.” 

Industry groups including the Motion Picture Association of America and the Streaming Innovation Alliance opposed the bill, claiming streamers were already working to address the issue, and noting that they have to deal with a variety of output devices, including TVs, tablets, and phones.