EPIC FAILS in Application Development Security practice processes, training, implementation, and incident response
AdaptHealth says attackers sweet-talked their way into cloud systems and stole patient data
AdaptHealth says attackers used social engineering to breach its systems and steal sensitive patient data, including passwords associated with insurance billing. The attack targeted an unwitting third-party contractor, through which the cybercriminals gained entry to the company's cloud environment, where they accessed business applications holding sensitive data. According to a 2024 annual report, it serves more than 4.2 million patients across all 50 US states.
The company's response included disabling the contractor's user account, resetting credentials, and implementing additional access controls. It believes the attack is now contained.
[rG: Roles based “Access Controls” is the key here. Administrative accounts should not provide access to sensitive information data. Human and service accounts need to have restricted access to prevent lateral exploits. Also, strong encryption of data stores, using restricted fields access prevents data copy/exfiltration.]
Accenture admits to 'isolated matter' after crook tries to flog alleged 35GB haul
Accenture has confirmed an "isolated matter" after a cybercriminal put up for sale what they allege is 35GB of the consulting giant's internal data, including source code, cryptographic keys, and cloud credentials.
The same threat actor previously claimed responsibility for another Accenture-related incident in 2024, alleging that data connected to more than 32,000 current and former employees had been exposed through a third-party compromise. Whether that earlier claim was genuine remains unclear.
So far, Accenture is keeping the technical details to itself. Whether this was a single exposed repository or something with a longer tail is anyone's guess until the company decides to say more.
An unnamed US county – perhaps in Ohio – paid $1M extortion demand to cybercriminals
An extortion gang that claimed to have stolen more than 2 TB of data, but the county never received independently verifiable proof that the stolen files had been deleted - just the criminals' promise.
After listing the victim county on their name-and-shame blog, Kairos demanded $3 million. The county then responded: “We have reviewed the situation with our leadership and financial teams. As a small county with very limited resources, we simply do not have the ability to meet the amount you have proposed. That said, we understand the seriousness of the matter and want to work toward a resolution. The most we have been able to identify at this time is $100,000. We respectfully ask that you consider this offer.”
Kairos responded: “You are wasting our time with such offers.We cant accept it.Your files will be a great advertisement on our site and we understand what terrible consequences will await you. You cant hide the data leak. You have two more days to make us a favorable offer.”
Two days later, the county increased its offer to $255,000. Kairos reduced its demand to $2 million, and on June 9, 2025, the county proposed paying $430,000.
That same day, both parties settled on $1 million.
Kairos claimed to have gained initial access by bruteforcing their way into the network
A payment ban will backfire because it doesn't address the root cause of our national problem: widespread digital insecurity.
GitHub AI agent leaks private repos when asked nicely
Malicious prompters could easily trick GitHub agents into pulling data from private repositories and then leaking the information as a public comment for anyone to access.
The issue exists in GitHub’s Agentic Workflows, which allow an AI agent powered by Claude or GitHub Copilot to autonomously execute tasks in GitHub Actions.
The attacker simply hides the malicious commands in plain English in the issue body, and the agent will then post this data as a public comment on the issue in the public repository.
To exploit this vulnerability, the attacker needed no coding skills, access, or credentials. All that was needed was to open an issue in a public repository belonging to an organization that uses GitHub’s Agentic Workflow setup and wait.
The vulnerability can’t be completely fixed in code. So the Noma researchers proposed documentation instead – but that didn’t happen, either. The proposed fix was a documentation callout encouraging users to adopt different strategies to their API key sharing between their repos.
An autonomous agent should not be a risk for silent data exfiltration and secrets exposure. Before a security team gives a pass to any autonomous agent, they need to ensure they understand all possible connections, access and paths, potential blast radius of the agent's access, and permissions. You can't protect what you can't see and control.
[rG SSDLC: Importance for solution CI/CD pipeline security design reviews to require complete and accurate solution design diagrams.]
GitHub Copilot: Sorry Dave, I can't do that harmful thing - unless you ask me in code
GitHub Copilot refuses harmful prompts almost always if asked in chat - like, "how to fool a breathalyzer test" or "smuggle bulk cash out of the US" - but then will write them in code 100 percent of the time if the prompt is broken into smaller steps and distributed across multiple stages of a software development workflow.
The results suggest that prompt-level safety evaluations aren’t sufficient for testing coding-agent safety. Coding-agent safety cannot be measured only by asking: Does the model refuse this malicious prompt?
They suggest developing model-safety benchmarks that exist inside live agentic workflows that not only score the final output, but also the “trajectory of turns, intermediate files, generated examples, and artifacts that led to it.”
Additionally, coding-agent developers should build in guardrails that examine the files, scripts, and data structures an agent writes - not just the chat reply - and reason over the entire session trajectory.
Your next insider threat doesn’t have a badge. It has an API token
Stop worrying about prompt injection; your next major breach will be caused by an authorized AI agent doing exactly what you permitted it to do.
The problem is that each security check only looks at one step at a time. Is this read okay? Yes. Is this summary okay? Yes. Is this email okay? Yes. Each step passes. But nobody is watching the combination of all three steps together. The security tools designed for human-driven workflows assumed a person would be doing this manually, one thing at a time. However, the AI agent bundles it all into a single automated sequence, and that bundling slips through gaps between the checks.
Patch for Windows Defender 0-day could allow attackers to fill hard disk
A patch Microsoft released on Wednesday to fix a zero-day vulnerability in its Defender security engine may cause Windows machines to write files large enough to completely consume available disk space.
RoguePlanet, tracked as CVE-2026-50656, came to public notice in June when NightmareEclipse, the pseudonymous name used by a researcher, disclosed it along with code for exploiting it. The vulnerability allows remote attackers to gain administrative control of Windows 10 and Windows 11 machines, even when real-time protection has been disabled. Over the past few months, the anonymous researcher has published a handful of other zero-days that have sent Microsoft scrambling to develop patches.
Attested TLS: the handshake that can't prove who's on the other end
Confidential computing rests on a mechanism called remote attestation, in which a server cryptographically proves to a client that it is running inside a genuine, unmodified Trusted Execution Environment (TEE) before any sensitive data changes hands.
Researchers at TU Dresden, spent the past two years formally verifying whether attested TLS, actually does what it claims. Using ProVerif, a tool for the symbolic security analysis of protocols, they discovered that it largely does not.
A connection intended for one server can be silently redirected to a different, compromised machine running identical software, anywhere in the world, without the client ever knowing. The intended server has done nothing wrong. The attacker simply exploits the fact that the protocol checks the software's integrity, not its location.
Another research team examined what the industry calls intra-handshake attestation, where evidence is generated during the TLS handshake itself, and tests seven different ways of cryptographically binding that evidence to the underlying connection. None of them prevent relay attacks, in which a client verifies the evidence of a genuine, trustworthy AI agent or server but ends up encrypting its traffic to an entirely different, malicious one. The starting assumption in all of this is that the hardware itself can be trusted.
[rG: Looks like we’ll see another TLS version coming up in a few years.]
Secret Claude tracker shocks users after Anthropic’s anti-surveillance stance
The AI firm was using “prompt steganography” to hide code that tracks Chinese users “in plain sight.” This code wasn’t malicious, but it was sending information to Anthropic that most users wouldn’t detect, relying on shorthand markers to quietly flag users’ timezone, proxy, and potential connection to Chinese AI labs that Anthropic has accused of distillation attacks.
Alibaba banned its employees from using Claude Code. “As Claude Code was recently discovered to carry back-door risks, after comprehensive evaluation, Claude Code has now been added to a list of high-risk software with security vulnerabilities.”
Privacy advocates were not happy with the explanation, though, warning that the code is evidence that Anthropic is willing to cross lines to surveil users. That’s perhaps especially surprising, considering that Anthropic riled the Trump administration by refusing to allow the US government to use Claude to surveil US users.
Anthropic confirmed that the tracker was added to Claude Code as an “experiment” in March. The code was meant to prevent account abuse from unauthorized resellers and protect against distillation. Unauthorized retailers have sold access to free models for $1 a month, and pro subscriptions that can cost $100 monthly sell for as little as $12.
Anthropic has been meaning to take this down for a while because engineers have “landed stronger mitigations since then.”
US firms like Anthropic are taking “increasingly aggressive measures” to block Chinese AI firms from copying their models.
In the past year, Chinese firms have “consistently matched” US firms’ model capabilities within months. Most recently, a new, free AI model from Chinese company Zhipu AI was better at finding computer vulnerabilities than Anthropic’s Claude Opus 4.8 model, which was released in May.
Lawmakers probe growing use of Chinese AI models in U.S. companies
Rising adoption of China-built AI models has led to growing calls from U.S. lawmakers for strategies to combat the trend, including via an ongoing investigation from two U.S. House Committees.
An initial step in the probe was for the chairmen of those committees to send letters to Cursor and Airbnb, over their “use of or exposure to these risks” through AI developed in China.
Chinese models are gaining traction among U.S. firms as they close the performance gap with American rivals while being cheaper to use. While some government departments have banned the usage of Chinese AI models including DeepSeek, adoption of them by U.S. companies is not prohibited.
Cursor, which will be acquired by Elon Musk’s SpaceX for $60 billion, built its Composer 2 model using Chinese AI model Kimi, which was developed by Moonshot AI.
One approach could be procurement requirements that discourage companies that want to do business with the government from using Chinese AI models
Enterprise AI still smarting from leaping before looking
The majority of companies that deploy AI systems end up shooting themselves in the foot with security.
78% of enterprises report "experiencing AI-related security incidents or identifying AI-related vulnerabilities. Among respondents,
27.7% experienced one incident,
21.9% experienced multiple incidents, and
28.4% had no incidents but identified vulnerabilities.
While 90% of organizations surveyed have discussed AI governance at the board level, just 50% have dedicated AI governance budgets and formal governance programs. This allows operational blind spots to persist.
Just 53% of respondents said their organization could trace AI decisions back to the models and source data that produced those results.
DigiCert attributes the survey's findings to lack of AI governance. "We wouldn’t allow an employee to operate without a verified identity. AI agents should be no different.”
Spacelift found 93% of organizations experienced AI-caused infrastructure incidents while only 19% had a governance plan in place.
There are several initiatives underway to establish identifiers for bots, such as Private Access Control Tokens (PACTs), Estonia's digital IDs for agents, and Microsoft's Agent ID. But bot badging infrastructure remains a work-in-progress, leaving AI agents to run amok in many organizations.
[rG: Governance plan and program is the easy part. But technical monitoring, enforcement, and security verification are key for incident prevention and response effectiveness.]
Gartner warned that more than 40% of agentic AI projects could be canceled by 2027. The issue is not just model capability. It’s governance, data access, ownership and ROI.
49% of security decision-makers flagged agentic AI as a concern. Read closely, that's a warning about what happens when agents get access and authority before anyone has sorted out who's accountable.
Before approving the next agent pilot, an executive should be able to get three answers in plain language.
1. What is the written success metric, and who agreed to it?
2. What data and tools does the agent actually need to reach, and does it have that access today?
3. When it fails, who notices, who owns the outcome, and how fast can someone roll it back?
[rG: #1 is product management financial analysis. #2 should be part of technical solutions’ Security Design Reviews analysis. #3 should be part of business operational analysis.]
HACKING
Researchers Claim First Fully Agentic Ransomware: JadePuffer
The campaign targeted an internet-facing Langflow instance by exploiting CVE-2025-3248. It then ran “an adaptive and fully automated campaign” which resulted in “a destructive database-extortion playbook against the victim's production database server.
Attack capabilities were delivered by an agent rather than a human-driven toolkit, the Sysdig research claimed, and the AI was capable of working autonomously retrying failed steps within refined parameters.
In one sequence, it went from a failed login to a working fix in 31 second.
Windows is watching: Anti-piracy tool fingers Scattered Spider suspect
The US Justice Department's complaint against Peter Stokes for alleged involvement in the Scattered Spider hacking group offers a reminder that it's difficult to hide online activity from Microsoft's operating system (or any other).
The complaint, arrest, and extradition of Stokes relied in part on a Microsoft Windows Global Device Identifier (GDID), among other telemetry records, to link online activity to the suspect.
A Global Device Identifier in the Windows ecosystem is a persistent, device-level identifier designed to uniquely identify an installation of a Windows operating system on a device, either a physical device (e.g., a mobile phone or laptop) or virtual machine, across certain Microsoft services and scenarios
Investigators obtained IP address records from ngrok and the VPN provider and then obtained records from Microsoft that matched the time when that ngrok account had been set up on a Windows machine through a specific GDID.
Apple maintains similar identifiers, including a hardware UUID and a DSID (Destination Signaling Identifier) tied to iCloud, among others.
Linux also supports a machine-id. And when presented with a lawful demand for information, most service providers will cooperate and provide whatever information they store.
“HalluSquatting” weaponizes LLMs’ inability to say “I don’t know.”
AI prompt injection has quickly become the top threat. Large language models are inherently unable to distinguish between legitimate instructions provided by users and malicious ones sneaked into emails, source code, and other third-party content the models are processing.
A new attack the researchers have named HalluSquatting has the potential to assemble massive botnets, perform large-scale DDoSes, and infect devices at scale, a first for prompt-injection attacks. The attack works against AI coding assistants and agents, including Cursor, Cursor CLI, Gemini CLI, Windsurf, GitHub Copilot, Cline, OpenClaw, ZeroClaw, and NanoClaw, which are all susceptible.
When a developer instructs a coding agent to clone a popular new repository, the LLM hallucinates its correct location 85% of the time. When cloning a trending “skill,” a form of instruction, script, or resource that gives agents specialized capabilities and domain expertise, hallucinations can occur 100% of the time. HalluSquatting focuses on trending resources because they aren’t included in the LLM training.
Attacks like HalluSquatting provide a potent reminder that some of the efficiencies are exaggerated since, at the end of the day, users must double-check details such as the location for each resource incorporated into a project. It also provides a cautionary lesson on the unintended and potentially dire outcomes that can result when people rely too heavily on AI assistants.
GuardFall: Artificial IntelligenceDecades-Old Bash Tricks Expose AI Coding Agents to Supply Chain Attacks
If an engineer uses a vulnerable agent to read a poisoned README or Makefile from a malicious repository, the agent can be tricked into silently executing commands that exfiltrate AWS credentials or wipe whole dev environments – especially in CI pipelines where ‘auto-yes’ modes are default.”
The ‘gap’ is a failure to guard the agent against the decades old Bash shell tricks, such as quote removal and $IFS spacing. Since these agents run with a developer’s full account authority, this can radiate into a major supply chain risk.
GuardFall: bypasses against pattern-based shell guards in agentic coding tools, where Bash unwinds the obfuscation after the guard has let the command through.
Adversa examined the most popular open-source coding agents and computer use agents as of May 2026, based on GitHub star count and community activity. Not all of the agents failed, but the bottom line is that only 1 of the 11 tested agents blocked all of the tricks.
Russia hacks doorbell cameras to spy on Nato bases
The Russian operation had targeted cameras pointing towards military transport routes in the hope of identifying what weapons were being sent to Ukraine.
The Dutch intelligence agencies did not specify the type of cameras hacked, the doorbell systems are frequently used by people to monitor their property from mobile phones. Hackers then use readily available apps to scan for devices that might be accessible.
A precisely timed laser pulse, aimed at the chip inside a Tangem crypto wallet card, can reset the card's password to anything the attacker picks. No old password. No backup card. Once it is reset, whoever did it controls the wallet and can move the coins out.
The company called this a lab-only physical method that works against secure element chips in general, not something unique to its cards.
In 2024, engineers rescued Voyager 1 after it spent months transmitting gibberish, remotely rewriting and relocating pieces of 46-year-old code around a failed memory chip from more than 24 billion kilometres away
Engineers had traced the trouble to the Flight Data Subsystem, one of the spacecraft’s three onboard computers, and the one that gathers and packages information before it is transmitted.
A single chip in that computer’s memory had failed. A stuck bit had knocked out roughly three per cent of the subsystem’s memory, and, crucially, that memory held part of the software’s own code. Around 256 words of storage were simply gone.
Voyager’s computers are designs from the early 1970s, with a total memory a tiny fraction of what sits in an ordinary phone today. Most of the engineers who built them have long since retired, so the team fixing the fault had to work from decades-old documents and manuals.
So the team took the affected code apart. They broke it into sections, found room for each section in a different corner of the memory, and rewrote the internal cross-references so that the scattered pieces would still call on one another and run as a single working program.
A command sent to Voyager 1 has to cross about 24 billion kilometres, which takes light, and therefore radio, roughly 22.5 hours. The reply takes just as long, so a single question and answer is close to a two-day exchange.
APPSEC, DEVSECOPS, DEV
Wars are blurring lines between corporate and national security
Companies now need to protect the data networks that serve as gateways to critical infrastructure. Hackers increasingly target not just computer files to steal information but also systems managing vital functions like building access and factory control, remotely causing physical damage or enabling espionage.
The North Atlantic Treaty Organization’s 32 countries last year agreed that as part of a pact to spend 5% of economic output on defense and security, 1.5% would go to military-adjacent needs including protecting critical infrastructure and networks.
The scale of the threat is prompting calls to elevate responsibility for corporate security to the C-suite. Executives and business-risk experts in March published a call to action in the Harvard Business Review titled “The Case for Hiring a Chief Resilience Officer.”
Companies say they need greater clarity from governments on what protections they will provide and subsidies to help them defend privately owned assets that provide a public good.
Most governments don’t provide incentives for companies to invest more than the minimum legal resilience requirements. That will likely make it an area for competition among businesses, said Heit at Vodafone, and the companies that are able to invest will win out by touting their resiliency as a benefit to clients.
[rG: Best practice software security design includes business continuity resiliency and incident response automatic failovers – requiring whole dependency chain protections and geographically separated multi-instance processing.]
UK Government's cyber pledge lands 60 signatories, including M&S and, somehow, Capita
The UK government's new Cyber Resilience Pledge signatories commit to treating cybersecurity as a board-level responsibility, signing up to the National Cyber Security Centre's Early Warning service, and encouraging suppliers to achieve Cyber Essentials certification or an equivalent baseline.
VENDORS & PLATFORMS
After spooking Trump into safety testing, Anthropic AI models get global release
The US has lifted export curbs on Anthropic’s newest Claude models, Fable 5 and Mythos 5, about three weeks after the Trump administration flagged the models as national security risks.
Fable 5 will be available globally, and US organizations have had access restored to Mythos 5 since June 26. Anthropic said it is now working with the government to expand Mythos access to a “broader set of domestic and international partners in the Glasswing program.” That program allows cybersecurity researchers at trusted companies to access Mythos for defensive purposes.
OpenAI ChatGPT Work to do your work for you and with you
Rebranded Codex promises independent workflows that can run “for hours if needed.”
ChatGPT Work also integrates Scheduled Tasks, a souped-up version of cron jobs. These tasks can keep going when you’re away from your desk and can be monitored from your phone. It can connect to common workplace management tools like Slack, Microsoft Teams, Google Drive, and SharePoint through various custom-built plugins. The desktop version of the new tool can also access and modify your desktop files or use a built-in browser to access online resources. OpenAI says an updated ChatGPT Chrome extension will also let you perform web-native tasks without switching out of the browser (OpenAI says its dedicated Atlas web browser is now being sunset, less than nine months after its launch).
Tool promises to make lazy academics' AI-written papers sound more human
AI humanizers are nothing new - take a cursory look online and you’ll find that companies pushing AI that helps AI writing sound less like AI wrote it are a dime a dozen. None, says the team behind Academic Humanizer, have been specifically tailored to papers and grant proposals.
The problem is that AI-assisted drafts come out generic and verbose, with … inflated phrasing, and over-long sentences. They also drift from the author's own voice and lose the precision scholarship depends on.
Academic Humanizer is, at its core, just a Claude skill, so it’s using more AI to correct the problems introduced into the academic writing space by AI. To further make its generative utterances more lifelike, Academic Humanizer can be pointed directly at its user’s prior work so that it’ll sound more like the “author” of the new paper instead of a bot.
IBM Bob Adds Multi-Agent AI and Cost Controls as Token Bills Become Boardroom Issue
IBM expanded its Bob agentic development platform with formal multi-agent orchestration, isolated subagents, and a built-in cost analytics dashboard called Bobalytics — a set of additions timed precisely to the moment when unpredictable AI token costs have become a procurement crisis for enterprise engineering teams.
The announcement lands one day after a Microsoft evaluation found that upgrading from one generation of AI model to the next multiplied token consumption by a factor of 10 to 12 in complex agentic coding scenarios — with one test run consuming 69 million tokens in a single task. In parallel, GitHub's June 1 shift to token-based billing for its 4.7 million paid Copilot subscribers generated projections of cost increases between 10x and 50x for developers running agentic sessions. The economics of enterprise AI tooling are, in the words of the engineering teams now dealing with them, no longer a technical problem but a budget problem.
The AI coding assistant market registered $12.8 billion in revenue in 2026, with GitHub Copilot at 4.7 million paid subscribers, Cursor at $2 billion in annual recurring revenue, and Claude Code at the top of developer satisfaction rankings in a JetBrains survey. But the market structure matters for understanding what IBM Bob is competing against.
The dominant AI coding tools accelerate individual developer velocity.
Suspecting AI cheating, Ivy League prof ordered an in-person final; scores fell 50%
Serrano decided that his spring 2026 section of the quite difficult ECON 1170 would allow take-home exams for both the midterm and the final. Suddenly, the course received an influx of students.
The results of the midterm exam were extraordinary, with an average score of 96 out of 100. Forty students scored a perfect 100. Historically the average grade in the midterm of this course has ranged between 65 and 80 [percent], and this exam was harder than the exams I wrote in the past, because… take-home is an opportunity to challenge the class a little bit more, given that you’re giving the students unlimited time.
A suspicious Serrano decided that he would make the final exam in-person; he would see if students did similarly well on it. 18 students suddenly dropped the course, while 9 others didn’t even attend the final exam. Of those 27 students, “22 had scored a perfect 100 in the midterm exam.” Among those who took the test, the average score plunged from 96 to 48.
The Big Con of Agentic AI
In 2023, the economists Mariana Mazzucato and Rosie Collington published The Big Con, arguing that a confluence of incentive structures, information asymmetries, and institutional pressures had allowed the management consulting industry to extract returns in excess of the value it creates. Surely, sophisticated clients would simply stop hiring expensive firms that did not deliver? Not quite, since a client who outsources a strategic function over many years may no longer possess the internal expertise to evaluate whether the advice it receives is sound, which is a key dynamic that perpetuates the engagement. The parallel with today’s AI vendors is noteworthy: deep integration, high costs, and a customer base that progressively grows less capable of evaluating what it is being sold.
The mechanism at work is something akin to the dark side of what management scholars call “unlearning by not doing.” The less an organization performs a function internally, the less it knows how to do it; the less it knows, the more it needs outside help; and the more it pays for outside help, the less it builds the knowledge that would eventually make outside help unnecessary.
The risk is especially acute for strategically central functions, because once those functions migrate outside, rebuilding them internally may take years.
LEGAL & REGULATORY
Ryuk ransomware member pleads guilty in the US, faces 15 years in prison
According to court documents, Vardanyan helped deploy Ryuk ransomware on the networks of multiple U.S. organizations between November 2019 and April 2020 after illegally accessing their systems.
In one attack, Vardanyan and his co-conspirators breached a Michigan company that paid 200 BTC (worth more than $1.1 million at the time). Two other attacks the prosecutors noted include a technology company in Wilsonville, Oregon, and a school in Texas.
The DoJ says that Vardanyan and his co-conspirators received about 1,610 bitcoins in ransom payments, valued at around $15 million at the time. It is estimated that at its peak, the Ryuk ransomware gang hacked around 20 organizations every week and made more than $150 million.
The ransomware operator faces a maximum sentence of 15 years in prison for two separate charges, as well as fines of $250,000 each. As part of his plea agreement, Vardanyan has agreed to pay more than $1.1 million in restitution.
Increased drone surveillance of illegal July 4th fireworks led to $100K fine
The Sacramento Fire Department issued 70 citations on July 4, totaling $300,000 in fines. But the department was also looking to assign additional citations on July 2 and July 3. The department had previously issued a $1 million fine to a single individual in 2025.
The Anaheim Police Department used drones to help issue 40 citations and confiscate 2,500 pounds of illegal fireworks.
The resulting $100,000 fine came from counting the number of fireworks and possibly other factors. Sacramento County assigns fines for illegal fireworks starting at $1,000 per device and as high as $10,000 per device used near sensitive areas such as schools or parks. But felony criminal charges punishable by imprisonment are also possible if illegal fireworks cause a fire that leads to property damage or bodily injury.
Supreme Court allows Texas to require age verification for mobile apps
The Supreme Court on Monday allowed Texas to enforce a law that requires mobile app stores to verify the age of users and obtain parental consent for minors attempting to install programs on their phones.
The Supreme Court’s emergency docket decision doesn’t resolve the case but rather will allow Texas to enforce the law while the litigation continues to play out.
Texas’ law, enacted last year in response to concerns about the online safety of minors, was opposed by the computer industry and a group of students who argued that it violated the First Amendment. Other states have considered similar laws amid a push to tighten online regulations for young people.
“A minor child who downloads a software application from an app store agrees to contractual terms of service, including whether the child’s location will be tracked, whether the child’s privacy will be protected, whether information from the child’s phone can be sold by the developer, and whether the child waives the right to sue.”
The European Union is ramping up pressure on Meta to make big changes to Facebook and Instagram after the European Commission preliminarily found that features like autoplay, infinite scroll, and highly personalized content recommendations were addictive.
On Thursday, the EC said its investigation indicated that “Meta did not adequately assess the risks of its addictive design on the physical and mental wellbeing of users, including minors and vulnerable adults.”
[rG Next Up: Other social media platforms, video platforms, virtual worlds, e-games, ...]
And Now For Something Completely Different …
Elite freedivers can hold their breath for more than 11 minutes on a single breath of air - and one oxygen-assisted record stands at nearly 30. Most untrained people tap out well before ninety seconds
We think of astronauts as floating in zero gravity, but the International Space Station is still deep inside Earth’s pull, where gravity is nearly 90% as strong as it is on the ground. What looks like weightlessness is really the station and its crew fall
What looks like weightlessness is something else entirely: the station and everyone aboard it are falling around the planet at about 28,000 kilometres per hour.
Isaac Newton captured the idea centuries ago with a thought experiment about a cannon on a very high mountain. Fire the cannonball gently and it arcs down to the ground nearby. Fire it harder and it lands farther away, its path curving as the ground curves beneath it. Fire it hard enough and the ground curves away just as fast as the ball falls, so it keeps missing the Earth and comes all the way around. That is an orbit: a projectile falling continuously toward the planet while moving sideways fast enough to keep overshooting it. The space station is doing exactly this.
Earth reaches its farthest point from the Sun on July 6, in the middle of northern summer, and the backward timing is the clearest sign the seasons don't come from distance
At aphelion this year Earth will be 94,502,961 miles from the Sun, or about 152.1 million kilometers. Six months from now, in early January, it will swing in to its closest point, called perihelion, at roughly 91.4 million miles. The gap between the two is a little over three million miles, around five million kilometers. The orbit is very nearly a circle, just slightly off. Across a full year our distance from the Sun changes by only a little over three percent.
Aphelion arrives about two weeks after the June solstice, and perihelion about two weeks after the December solstice. It looks like a pattern, though it is closer to a coincidence. The dates also wander. Because the shape and orientation of Earth’s orbit shift slowly under the pull of the Moon and the other planets, the timing drifts by roughly a day every 58 years, and can move a day or two from one year to the next. By the same reckoning, the December solstice fell on the very day of perihelion back in 1246, and thousands of years from now perihelion will have slid around to the March equinox.
China recovered its first reusable rocket and showed a new way to do it
The Long March 10B booster descended from space and guided itself into a four-legged frame affixed to an offshore vessel. Tensioned cables stretched over the ship in a grid pattern captured the rocket as it shut down its landing engines, leaving the smoldering booster hanging in midair. Catching the rocket in this way reduces the effect of reuse on payload capacity. The Long March 10B doesn’t have to carry the extra mass of landing legs, and recovering it downrange reduces how much fuel the rocket must consume during its descent.
