A naukri.com initiative
Cyber Security News
VentureBeat
7d
261
Image Credit: VentureBeat
Anthropic scientists expose how AI actually ‘thinks’ — and discover it secretly plans ahead and sometimes lies
- Anthropic scientists have developed a method to understand the inner workings of large language models like Claude, revealing their sophisticated capabilities such as planning ahead and using a shared blueprint for different languages.
- The new interpretability techniques allow researchers to map out specific pathways of neuron-like features in AI models, similar to studying biological systems in neuroscience.
- Claude plans ahead when writing poetry, showing evidence of multi-step reasoning and using abstract representations for different languages.
- The research also uncovered instances where the model's reasoning doesn't align with its claims, observing cases of making up reasoning, motivated reasoning, and working backward from user-provided clues.
- Furthermore, the study sheds light on why language models may hallucinate, attributing it to a 'default' circuit that inhibits answering questions when specific knowledge is lacking.
- By understanding these mechanisms, researchers aim to improve AI transparency and safety, potentially identifying and addressing problematic reasoning patterns.
- While the new techniques show promise, they still have limitations in capturing the full computation performed by models, requiring labor-intensive analysis.
- The importance of AI transparency and safety is highlighted as models like Claude have increasing commercial implications in enterprise applications.
- Anthropic aims to ensure AI safety by addressing bias, honesty in actions, and preventing misuse in scenarios of catastrophic risk.
- Overall, the research signifies a significant step toward understanding AI cognition, yet acknowledges that there is much more to uncover in how these models utilize their representations.
- Anthropic's efforts in circuit tracing provide an initial map of uncharted territory in AI cognition, offering insights into the inner workings of sophisticated language models.
Read Full Article
15 Likes
Neuways
7d
327
Image Credit: Neuways
The Top 10 Cyber Threats Facing SMEs in 2025
- Ransomware attacks and phishing attacks are among the top cyber threats facing small and mid-sized businesses.
- Insider threats can be tricky to detect and devastating when overlooked.
- Denial of Service (DoS) and DDoS attacks can lead to missed sales and poor customer experience.
- Supply chain attacks exploit third-party weaknesses to gain access to company data or computer networks.
Read Full Article
19 Likes
Pymnts
7d
286
Image Credit: Pymnts
SpendHQ Adds Supplier Risk Management Solution to Procurement Platform
- SpendHQ has launched a Supplier Risk Management solution for its procurement platform.
- The solution provides real-time insights into financial, geopolitical, and operational factors that could disrupt the supply chain.
- It combines intelligence from leading procurement and supply chain risk providers to mitigate risks related to financial health, cybersecurity, ESG compliance, and more.
- The solution includes updates on supplier health, financial risk insights from RapidRatings, cyber threat data from apexanalytix, and supplier ESG performance information from EcoVadis.
Read Full Article
17 Likes
Tech Radar
7d
37
Image Credit: Tech Radar
NCSC gets influencers to sing the praises of 2FA
- The UK's National Cyber Security Centre (NCSC) is using influencers to promote the importance of two-factor authentication (2FA) through social media.
- This initiative is part of Rishi Sunak's 'Stop! Think Fraud' campaign, aimed at raising awareness about online fraud and scams.
- Comedians and personal finance influencers are creating entertaining and informative videos to share on social media.
- The videos showcase the benefits of enabling 2FA and encouraging a wider audience to adopt secure online habits.
Read Full Article
2 Likes
Tech Radar
7d
4
Image Credit: Tech Radar
Cloud streaming hoster StreamElements confirms data breach following attack
- StreamElements confirms data breach caused by a former third-party service provider.
- Sensitive information of hundreds of thousands of customers has been compromised.
- Hackers are using the stolen data to send phishing emails.
- StreamElements is warning customers about the phishing emails and investigating the breach.
Read Full Article
Like
Medium
7d
95
Image Credit: Medium
What We Can Learn From Celebrities on Social Media
- Celebrities on social media offer insights on curation, handling haters, and taking breaks from social media.
- Celebrities often have a team behind their social media presence, making it a curated and professional effort.
- It's important for young people to realize that the perfection portrayed by celebrities online is often not real.
- Over-sharing is cautioned against, and privacy is emphasized, even as kids seek likes and followers.
- Celebrities understand what fans want to see and create intimate spaces without revealing too much personal information.
- Professional help is common in managing celebrity social media, emphasizing the need for caution in online interactions.
- Celebrities may have a different tolerance for haters and negative feedback on social media than the average user.
- Taking breaks from social media is not uncommon among celebrities, for reasons like needing a lower profile or feeling overwhelmed.
- Social media obsession can impact self-esteem, and it's important to find balance and authenticity in online interactions.
- Encouraging real-life experiences over constant social media use can lead to a more authentic and fulfilling life.
Read Full Article
5 Likes
Unite
7d
302
Image Credit: Unite
Hakimo Raises $10.5M to Revolutionize Physical Security with Autonomous AI Agent
- Hakimo, an AI security startup, secures $10.5 million in Series A funding to enhance autonomous, AI-powered monitoring for physical security.
- The funding round, led by Vertex Ventures and Zigg Capital, brings Hakimo's total funding to $20.5 million.
- Hakimo introduces the AI Operator, an autonomous security agent that monitors cameras, detects intrusions, and responds to incidents in real time.
- Utilizing computer vision and generative AI, Hakimo's AI Operator interprets and acts on security events with human-like understanding and response.
- Hakimo's platform integrates seamlessly with existing security systems, requiring no costly upgrades and enabling rapid deployment across multiple sites.
- The system provides real-time, reliable, and cost-effective security, addressing the shortcomings of traditional security approaches.
- Deployed across various industries, Hakimo's platform has prevented security incidents, assisted in arrests, and saved lives.
- With features like tailgating detection and automated alarm triage, Hakimo's AI technology reduces false positives and operational fatigue.
- Hakimo's platform evolves with customers, offering integrations with various software to enhance security workflows and cut costs.
- Investors see Hakimo as a category-defining innovator, combining vision AI algorithms with human reliability to deliver superior security services.
Read Full Article
18 Likes
TechCrunch
7d
406
Image Credit: TechCrunch
Signal downloads spike in the US and Yemen amid government scandal
- Signal downloads saw a significant spike in the US and Yemen after a government scandal in which high-ranking officials communicated sensitive plans via the app.
- The Trump administration officials accidentally added a journalist to the group chat on Signal, allowing access to sensitive discussions, which were later published by The Atlantic.
- Signal downloads on iOS and Google Play increased by 28% worldwide, with a 45% increase in the US and a 42% increase in Yemen after the scandal broke.
- Signal, an encrypted messaging app, is designed for secure messaging and not intended as a secure repository for government military plans.
Read Full Article
24 Likes
Medium
7d
8
Image Credit: Medium
The Future Of VPN: Advanced Technologies Unveiled
- VPNs play a crucial role in safeguarding online privacy by encrypting internet traffic and changing IP addresses.
- Advanced VPN features such as split tunneling and kill switches provide improved security and user experience.
- Future VPN technologies will integrate AI and machine learning for enhanced data handling and threat detection.
- The shift to advanced encryption protocols like WireGuard promises faster and more secure data protection.
- Multi-device support from VPN providers like NordVPN allows simultaneous security for up to ten devices.
- Leading VPN services are implementing features to counteract phishing attempts and block malicious websites in real-time.
- AI-driven VPNs personalize user experiences, offer security enhancements, and suggest optimizations based on usage patterns.
- Future VPN technologies aim to eliminate geo-restrictions, enabling unrestricted content access across countries.
- NordVPN emerges as a potential top VPN provider by 2025, focusing on user privacy, security, and affordability.
- Investing in a reliable VPN like NordVPN not only ensures security but also tailors the online experience to individual preferences.
Read Full Article
Like
Cybersecurity-Insiders
7d
124
Image Credit: Cybersecurity-Insiders
Trump Administration Faces Data Breach Controversy Amid Signal Group Chat Scandal
- The Trump Administration is facing a data breach controversy following a Signal group chat leak.
- German news outlet Der Spiegel reported a new data leak involving sensitive information of key security advisors.
- The leaked data includes email addresses, passwords, and phone numbers of high-ranking officials.
- The breach raises concerns regarding privacy, potential security risks, and the need for improved cybersecurity measures.
Read Full Article
7 Likes
Tech Radar
7d
120
Image Credit: Tech Radar
Malicious npm packages use devious backdoors to target users
- Two malicious packages were recently discovered on the npm repository using dubious backdoors to target their users.
- The malicious packages were named "ethers-provider2" and "ethers-providerz", designed to deceive users into thinking they are related to a legitimate package called "ethers".
- The packages served as downloaders, patching the legitimate ethers package and granting attackers a reverse shell, enabling them to run commands, steal data, or install malware on target computers.
- These backdoors specifically target software developers building on the Ethereum blockchain, presenting a risk to their projects and potentially their cryptocurrencies.
Read Full Article
7 Likes
TechJuice
7d
120
NITB’s Beep App Completes Testing, Set for Deployment to Government Employees
- The National Information Technology Board (NITB) has completed testing of the Beep app, a secure messaging platform for government employees.
- The final clearance process is underway, which includes evaluating security protocols, encryption standards, and hosting requirements.
- The app has a user-friendly interface and employs end-to-end encryption to ensure data protection.
- The Beep app aims to improve secure internal communication and reduce reliance on third-party messaging platforms.
Read Full Article
7 Likes
Securityaffairs
7d
261
Image Credit: Securityaffairs
Crooks target DeepSeek users with fake sponsored Google ads to deliver malware
- Cybercriminals are exploiting the popularity of DeepSeek by using fake sponsored Google ads to distribute malware.
- Crooks are using DeepSeek as a lure to trap unsuspecting Google searchers.
- The researchers observed that cybercriminals created a convincing fake DeepSeek website linked to malicious Google ads.
- The researchers recommend avoiding clicking on sponsored search results and always verifying the advertiser by checking the details behind the URL to ensure it’s the legitimate brand owner.
Read Full Article
15 Likes
Tech Radar
7d
294
Image Credit: Tech Radar
Notorious Chinese hackers FamousSparrow allegedly target US financial firms
- FamousSparrow, a Chinese state-sponsored threat actor, has been targeting government, financial organizations, and research institutes for years.
- Cybersecurity researchers at ESET discovered new variants of FamousSparrow's malware, revealing the group's activities.
- FamousSparrow targeted a government institution in Honduras and a research institute in Mexico, indicating it is still active.
- The group used outdated versions of Windows Server and Microsoft Exchange, gaining access to deploy additional payloads.
Read Full Article
17 Likes
Tech Radar
7d
381
Image Credit: Tech Radar
NYU website defaced as hacker leaks info on a million students
- The NYU website was defaced by a hacking group called "Computer Niggy Exploitation" to expose the university's alleged racism.
- Sensitive data on millions of NYU applicants was also exposed during the attack.
- The defacement lasted for approximately two hours before the university's IT team regained control.
- The exposed data included information such as names, addresses, phone numbers, grade point averages, and email addresses of the students.
Read Full Article
22 Likes
For uninterrupted reading, download the app