Tech

The Rise of Shadow AI and the Enterprise Response

The rapid adoption of AI agents like OpenClaw, formerly known as Clawdbot and Moltbot, is creating a significant challenge for IT and security departments. Employees, seeking increased productivity, are increasingly installing these tools on work machines despite documented security risks. This trend, dubbed “shadow AI,” is forcing organizations to rethink their security strategies.

OpenClaw: A Digital Master Key

Unlike traditional cloud-based LLMs, OpenClaw operates with potential root-level shell access to a user’s machine. This grants the agent the ability to execute commands with full system privileges, effectively acting as a “digital master key.” The lack of native sandboxing means sensitive data – SSH keys, API tokens, and internal records – are directly exposed. Runlayer CEO Andy Berman highlighted this risk, stating that a security engineer was able to gain full control of OpenClaw in under an hour using simple prompting.

The Threat of Prompt Injection

A primary technical threat is prompt injection – malicious instructions hidden within seemingly harmless communications like emails or documents. These instructions can hijack the agent’s logic, commanding it to exfiltrate sensitive data. For example, an email containing hidden system instructions could direct the agent to send customer data and API keys to an external server.

From BYOD to Shadow AI: A Historical Parallel

The current situation echoes the “Bring Your Own Device” (BYOD) trend of the past. Just as employees favored iPhones over corporate Blackberries due to superior technology, they are now adopting AI agents like OpenClaw for the “quality of life improvement” they offer. As Berman noted, the industry has moved past simply telling employees “no,” as they are already integrating these tools with Slack, Jira, and email, regardless of official policy.

Securing the Agentic Future: Runlayer’s Approach

Runlayer is addressing this challenge with a governance layer designed to transform unmanaged AI agents into secured corporate assets. Their ToolGuard technology provides real-time blocking with low latency, analyzing tool execution outputs to catch remote code execution patterns like “curl | bash” or destructive “rm -rf” commands. Internal benchmarks suggest this increases prompt injection resistance from 8.7% to 95%.

Discovery and Active Defense

Runlayer’s suite includes two key components: OpenClaw Watch, a detection mechanism for unmanaged servers, and Runlayer ToolGuard, the active enforcement engine that monitors every tool call. ToolGuard specifically targets credential exfiltration attempts, looking for the leakage of AWS keys, database credentials, and Slack tokens.

Licensing, Privacy, and the Vendor Model

Runlayer positions its solution as a proprietary commercial layer, SOC 2 and HIPAA certified, offering the legal and technical guarantees required by large organizations. The company emphasizes that it does not train on organizational data, and contracting with Runlayer is akin to contracting with a traditional security vendor.

Pricing and Deployment

Runlayer’s pricing is based on a platform fee, encouraging enterprise-wide adoption rather than per-user costs. The platform can be deployed in the cloud, within a private VPC, or on-premise, with all tool calls logged and auditable for integration with SIEM vendors like Datadog or Splunk.

Cultural Shift and the AI Transformation Team

Securing these tools can lead to a positive cultural shift within organizations. Gusto, for example, renamed its IT team to the “AI transformation team” after partnering with Runlayer. A customer at OpenDoor reported a significant “quality of life improvement” by being able to connect agents to sensitive systems without fear of compromise.

The Path Forward: Governance, Not Prohibition

Runlayer already secures AI for companies like Gusto, Instacart, Homebase, and AngelList, suggesting a future where AI is governed rather than banned. As model capabilities increase, the need for this infrastructure will only grow. Berman concluded that the goal is to enable a “governed, safe, and secure way to roll out AI.”

FAQ

• What is OpenClaw? OpenClaw is an open-source AI agent that runs locally and connects to messaging apps and a computer’s terminal.
• What is shadow AI? Shadow AI refers to the use of AI tools by employees without the knowledge or approval of IT and security departments.
• What is prompt injection? Prompt injection is a technique where malicious instructions are hidden within prompts to hijack an AI agent’s logic.
• What does Runlayer do? Runlayer provides a governance layer to secure AI agents like OpenClaw, offering discovery and active defense capabilities.

Pro Tip: Regularly review and update your organization’s security policies to address the evolving threat landscape of AI agents.

Learn more about securing your AI infrastructure. Explore Runlayer’s solutions.

The Space Sustainability Paradox: Can Satellites Save Earth Without Destroying Orbit?

Earth-observation satellites are increasingly vital for tracking progress towards the United Nations’ 17 Sustainable Development Goals (SDGs). These eyes in the sky provide crucial data on everything from land use and urban development to ecosystem health and disaster response. But, a growing body of research, including fresh work from The University of Manchester, highlights a critical challenge: the very tools we rely on to monitor and protect our planet could be jeopardizing the long-term sustainability of space itself.

The Rising Risk of Space Debris

The core of the issue lies in the increasing number of satellites orbiting Earth. As more are launched, the risk of collisions – and the creation of space debris – rises exponentially. Even small fragments of debris traveling at high speeds can cause significant damage to operational satellites, potentially creating a cascading effect known as the Kessler Syndrome, rendering certain orbits unusable.

Researchers at The University of Manchester have been investigating how to balance the need for comprehensive Earth observation with the imperative to maintain a safe orbital environment. Their work focuses on linking mission requirements – specifically image resolution and coverage – with satellite size and the total number of satellites deployed.

Size Matters: A Trade-off Between Coverage and Collision Risk

The study revealed a key trade-off: the size of each satellite directly impacts the probability of collisions. Larger satellites, while potentially offering broader coverage, also present a greater individual collision risk. Conversely, smaller satellites require a larger constellation to achieve the same level of coverage, increasing the overall number of objects in orbit and, the risk of collisions.

Interestingly, the research also found that higher orbits, while requiring fewer satellites for coverage, carry a greater individual collision risk due to the larger size of the satellites operating at those altitudes.

“Our research addresses what is described as a ‘space sustainability paradox’,” explains John Mackintosh, a PhD researcher at The University of Manchester and lead author of the study, “the risk that using satellites to solve environmental and social challenges on Earth could ultimately undermine the long-term sustainability of space itself.”

The University of Manchester’s Leading Role in SDG Research

The University of Manchester is recognized globally for its commitment to the UN SDGs. It’s the only university in the world to consistently rank in the top ten for social and environmental impact in the Times Higher Education Impact Rankings, currently holding the number two position worldwide (2025 rankings). The university also features in the top 10 of the 2025 QS World University Sustainability Rankings.

Beyond this new research, the University contributes significantly to SDG-related research, having produced 21,571 research publications across all 17 SDGs over the past five years – representing 4% of the UK’s total research output on these goals. Its research beacons – advanced materials, biotechnology, cancer, energy and global inequalities – are all focused on tackling global challenges.

What’s Next for Space Sustainability?

Addressing the space sustainability paradox requires a multi-faceted approach. This includes:

• Improved Space Traffic Management: Developing more sophisticated systems for tracking and managing satellites and debris.
• Debris Removal Technologies: Investing in technologies to actively remove existing debris from orbit.
• Sustainable Satellite Design: Designing satellites with finish-of-life plans that ensure responsible deorbiting.
• International Collaboration: Establishing international norms and regulations to govern space activities and promote responsible behavior.

The University of Manchester’s work underscores the urgency of these efforts. As our reliance on space-based technologies grows, ensuring the long-term sustainability of the orbital environment is paramount.

FAQ

Q: What are the UN Sustainable Development Goals?
A: The 17 SDGs are a collection of global goals set by the United Nations to address the world’s most pressing challenges, including poverty, hunger, climate change, and inequality.

Q: What is space debris?
A: Space debris consists of defunct satellites, fragments from collisions, and other man-made objects orbiting Earth. Even small pieces can pose a significant threat to operational spacecraft.

Q: What is the Kessler Syndrome?
A: The Kessler Syndrome is a hypothetical scenario in which the density of objects in low Earth orbit is so high that collisions between them become inevitable, creating a cascading effect that renders certain orbits unusable.

Q: How is The University of Manchester contributing to sustainability?
A: The University of Manchester is a global leader in sustainability research and education, ranking highly in international sustainability rankings and contributing significantly to research across all 17 SDGs.

Did you know? The University of Manchester is the only university in the world to rank in the top ten for social and environmental impact in every year of the Times Higher Education Impact Rankings.

Pro Tip: Stay informed about space sustainability initiatives by following organizations like the United Nations Office for Outer Space Affairs (UNOOSA).

What are your thoughts on the future of space sustainability? Share your comments below and let’s continue the conversation!

From Romantic Beginnings to Unexpected Trauma: How Life’s Curveballs Shape Relationships

The story of Camiel Kesbeke and Donna Bakker offers a poignant reminder that life rarely unfolds as planned, especially when it comes to love. Their relationship, blossoming from a teenage connection to a recent cohabitation, took an unforgettable turn during a first date marked by a harrowing event. This experience highlights the unpredictable nature of relationships and the strength required to navigate shared trauma.

The Date That Changed Everything

What began as a casual evening – drinks before Camiel headed to a party – quickly transformed into a scene neither he nor Donna could have anticipated. A traffic accident unfolding before their eyes, involving two women being thrown through the air, instantly shattered the romantic atmosphere. The sheer force of the impact, leaving a visible imprint on the car’s windshield, left a lasting impression on the couple.

The immediate aftermath was shocking. Beyond the trauma of witnessing the accident, an unexpected layer of awkwardness emerged when a familiar police officer arrived on the scene. Their budding romance was inadvertently “outed” as the officer questioned their presence together, turning a moment of distress into a slightly comical revelation.

Shared Trauma and Relationship Resilience

Experiencing a traumatic event early in a relationship can be a significant test. While not explicitly detailed in available sources, research suggests that shared experiences, even negative ones, can foster a deeper sense of connection, and intimacy. Navigating the emotional fallout of such an event requires open communication, empathy, and a willingness to support one another.

Camiel and Donna’s story underscores the importance of being able to lean on a partner during difficult times. Donna’s ability to inquire Camiel about his feelings, as noted in a recent interview, demonstrates a healthy dynamic where emotional vulnerability is encouraged. This is crucial for building a resilient relationship capable of weathering life’s storms.

The Power of Vulnerability in Modern Relationships

Camiel’s admission that he’s not naturally inclined to share his emotions, but does so with Donna, reflects a growing trend towards greater emotional openness in relationships. Traditionally, societal norms often discouraged men from expressing vulnerability. However, there’s a shift towards recognizing the benefits of emotional intelligence and authentic connection.

This trend is particularly evident among younger generations, who prioritize emotional intimacy and open communication in their partnerships. The ability to be vulnerable – to share fears, insecurities, and past experiences – is increasingly seen as a cornerstone of a healthy and fulfilling relationship.

Navigating Shared History and Individual Growth

Both Camiel and Donna share a common ground in their family experiences – having parents who are divorced. This shared understanding can create a strong bond, providing a space for mutual support and empathy. It suggests a capacity to navigate complex family dynamics and offer each other a safe haven.

Camiel’s experience on “Expeditie Robinson” seems to have spurred a desire for deeper connections. He recognized a difference between the superficial interactions he often had with male friends and the genuine emotional sharing he experienced during the show. This realization highlights the importance of seeking out relationships that foster authenticity and vulnerability.

Frequently Asked Questions

Q: Can witnessing a traumatic event strengthen a relationship?
A: While challenging, shared experiences, even traumatic ones, can foster a deeper connection if couples communicate openly and support each other.

Q: Is vulnerability important in a relationship?
A: Yes, vulnerability is crucial for building trust, intimacy, and a strong emotional bond.

Q: How can couples navigate differing levels of emotional expression?
A: Open communication, patience, and a willingness to understand each other’s emotional styles are key.

Q: What role does shared history play in a relationship?
A: Shared experiences, whether positive or negative, can create a sense of understanding and connection.

Did you know? Couples who regularly express gratitude towards each other report higher levels of relationship satisfaction.

Pro Tip: Schedule regular “check-ins” with your partner to discuss your feelings, needs, and concerns. This creates a safe space for open communication.

Want to learn more about building strong and resilient relationships? Explore our other articles on communication skills and emotional intelligence. Subscribe to our newsletter for weekly insights and relationship advice!

The Unexpected Resilience of Radiologists in the Age of AI

For years, the narrative surrounding artificial intelligence (AI) centered on job displacement, with roles involving data and image analysis – like radiology – often cited as prime examples. The reasoning was straightforward: AI excels at pattern recognition within images. Though, the predicted obsolescence of radiologists hasn’t materialized. Instead, radiology is emerging as a compelling case study in how AI can augment, rather than eliminate, human expertise.

Radiology’s Prominent Role in the AI Discussion

The intersection of AI and radiology has recently gained visibility on global stages, including discussions at the World Economic Forum in Davos and within a recent White House report on AI and the economy. This underscores the field’s significance in broader medical, political, and economic conversations.

Concerns about AI’s impact on radiology were amplified in 2016 when computer scientist Geoffrey Hinton, a pioneer in deep learning, suggested that aspiring radiologists reconsider their career paths. His prediction sparked debate and anxiety within the medical community.

How AI is Transforming the Radiologist’s Workflow

AI’s strength lies in its ability to rapidly process vast datasets. In radiology, where nearly all medical images are now digitized (X-rays, CT scans, MRIs), AI systems can efficiently analyze these images. This isn’t about replacing radiologists, but about enhancing their capabilities.

Currently, AI assists radiologists in several key areas:

• Prioritizing scans requiring immediate attention.
• Improving the quality of medical images.
• Providing summaries of complex medical reports.

Expert Perspectives on AI Collaboration

Dr. Po-Hao Chen, a diagnostic radiologist at Cleveland Clinic, highlights radiology’s unique position, noting the abundance of data available for AI research and training. Dr. Shadpour Demehri of Johns Hopkins Medicine emphasizes that AI doesn’t replace doctors; it streamlines their function, making it more efficient and meaningful. René Vidal, a professor at Penn Engineering at the University of Pennsylvania, points to AI’s potential to accelerate MRI scans by producing high-quality images with fewer measurements, increasing patient throughput.

Despite these advancements, the critical role of human doctors remains. Final diagnoses, patient examination, and the creation of official medical reports still require human judgment and expertise.

FDA Approvals and Job Market Trends

The implementation of AI in healthcare is subject to rigorous oversight. Medical AI tools require approval from the Food and Drug Administration (FDA), a process that can take up to eight years, encompassing development and clinical testing. As of February 2026, 1,401 of the 1,357 medical tools approved by the FDA with artificial intelligence belong to radiology.

Contrary to predictions of job losses, the radiology job market is experiencing growth. The Bureau of Labor Statistics projects a 5% increase in radiology jobs between 2024 and 2034, exceeding the overall average job growth rate of 3%.

Navigating the Risks and Limitations of AI in Radiology

Although AI offers significant benefits, potential risks must be addressed. A 2022 MIT study revealed that AI can accurately predict a person’s race from X-ray images, raising concerns about potential bias in medical decision-making.

Dr. Chen cautions against overreliance on AI, stressing that the system’s success hinges on human review of automated results. The most significant improvements arise from the collaborative synergy between machine and specialist.

Did you realize?

The FDA approval process for medical AI tools is extensive, ensuring patient safety and efficacy before widespread adoption.

The Future of AI and Radiology: A Symbiotic Relationship

Radiology demonstrates that technological advancements don’t necessarily equate to job losses. Instead, they often necessitate adaptation and the development of new skillsets. The future of work likely involves a partnership between professionals and technology, transforming how jobs are performed rather than eliminating them entirely.

FAQ

• Will AI replace radiologists? No, current evidence suggests AI will augment radiologists’ abilities, not replace them.
• How long does it take for AI tools to be approved for medical use? The FDA approval process can take around eight years.
• Is the radiology job market growing? Yes, the Bureau of Labor Statistics projects a 5% growth in radiology jobs between 2024 and 2034.
• Are there risks associated with using AI in radiology? Yes, potential risks include bias in AI algorithms and the danger of overreliance on automated results.

Desire to learn more about the latest advancements in medical imaging? Explore our articles on AI-powered diagnostics and the future of healthcare technology.

The Invisible Threat: How Spyware Like Predator is Silently Hijacking Your iPhone

Apple’s iPhone has long been lauded for its security features, but a new analysis reveals a disturbing trend: sophisticated spyware, like Intellexa’s Predator, is evolving to bypass even those protections. Researchers at Jamf have discovered Predator can now hide evidence of its surveillance, leaving users completely unaware their camera and microphone are being secretly accessed.

Predator’s Stealth Tactics: A Deep Dive

The core of Predator’s new tactic lies in manipulating iOS’s own security indicators. Since iOS 14, Apple displays green or orange dots on the status bar to signal when the camera or microphone is in use. Predator doesn’t exploit a vulnerability in iOS itself; instead, it leverages existing kernel-level access to intercept and suppress these indicators. Which means the spyware can stream audio and video without triggering any visible alerts.

Jamf’s research pinpointed a single “hook function” – ‘HiddenDot::setupHook()’ – within the SpringBoard application. This function is invoked whenever sensor activity changes. By intercepting this process, Predator prevents the status bar from displaying the camera or microphone icons. Essentially, the spyware silences the telltale signs of surveillance.

Beyond the Dots: How Predator Gains Access

Intellexa, the company behind Predator, has a history of exploiting zero-day vulnerabilities in Apple and Chrome, and employing “0-click infection mechanisms” – meaning the spyware can infect devices without any user interaction. The spyware has been linked to attacks targeting journalists, politicians, and activists globally, including a recent case in Angola where a journalist’s iPhone was compromised via a malicious WhatsApp link.

While the indicator suppression is a key component of Predator’s stealth, it’s not the only method employed. Researchers found that Predator uses ARM64 instruction pattern matching and Pointer Authentication Code (PAC) redirection to bypass camera permission checks, enabling access even without explicit user consent. For VoIP recordings, Predator relies on the same ‘HiddenDot’ function to conceal its activity.

The Broader Implications: A Growing Threat Landscape

The evolution of spyware like Predator highlights a concerning trend: commercial surveillance vendors are becoming increasingly sophisticated and adaptable. Despite sanctions imposed on Intellexa by the U.S. Government in 2024, the company continues to operate and develop new tools for surveillance. Google’s Threat Intelligence Group has documented Intellexa’s prolific use of zero-day exploits, accounting for 15 out of approximately 70 zero-days discovered by Google since 2021.

This isn’t just about technical capabilities; it’s about the potential for abuse. The targeting of journalists, as seen in Angola, underscores the threat to press freedom and civil society. The ability to silently monitor individuals raises serious privacy concerns and erodes trust in mobile technology.

Did you know? Predator doesn’t always rely on exploiting vulnerabilities. It can also gain access through social engineering tactics, like sending malicious links via messaging apps.

Detecting and Mitigating the Risk

Detecting Predator and similar spyware can be challenging, as they are designed to be stealthy. Jamf researchers note that technical analysis can reveal signs of malicious activity, such as unexpected memory mappings or exception ports in system processes. Though, this requires specialized expertise.

While a direct fix remains elusive, staying vigilant about security practices is crucial. Avoid clicking on suspicious links, keep your software up to date, and consider using mobile device management (MDM) solutions that can provide enhanced security monitoring.

FAQ: Predator Spyware

Q: What is Predator spyware?
A: Predator is commercial spyware developed by Intellexa, capable of remotely accessing and controlling mobile devices.

Q: How does Predator hide its activity?
A: Predator intercepts and suppresses the iOS camera and microphone recording indicators, preventing users from knowing they are being monitored.

Q: Is my iPhone safe?
A: While iPhones are generally secure, sophisticated spyware like Predator poses a risk. Staying vigilant and practicing good security habits is essential.

Q: What is a zero-day exploit?
A: A zero-day exploit is a vulnerability in software that is unknown to the vendor, giving attackers a window of opportunity to exploit it before a patch is available.

Pro Tip: Regularly review app permissions on your iPhone to ensure apps only have access to the data and features they need.

Learn more about mobile security threats and best practices by exploring resources from organizations like Amnesty International and Citizen Lab.

What are your thoughts on the increasing sophistication of mobile spyware? Share your concerns and experiences in the comments below.