1.145 casos · pág. 22/58
A US federal magistrate judge in San Francisco ordered former OpenAI engineer Abdulrahman Alfozan to remain in custody, citing flight risk. Prosecutors allege he planted a covert audio recording device, hidden in toilet paper behind a TV, in an executive conference room used for confidential meetings. OpenAI found the device in late Septe…
Canada's proposed C-22 bill, first put forward in 2025 and since partly revised, would require tech companies, including foreign ones, to introduce "technical capabilities" for law enforcement data access. It would also require retention of user metadata. Signal has said it would exit the Canadian market, and Psiphon VPN says it will leav…
Privacy firm Redact found that around 80 WhatsApp help and FAQ pages exposed hidden text instructions written for a Meta large language model support chatbot. The text, about 100 blocks in total, sat in the page source under a code block headed "HelpContentGalaxyLLMInclude". Some instructions told the chatbot not to recommend contacting a…
A Chainguard survey released at its Assemble London event on October 14 found that 66% of security leaders at companies with $1bn or more in revenue admit leaving high or critical vulnerabilities unpatched "at least sometimes". Among companies with under $250m in revenue, the figure was 27%. Larger firms cite business trade-offs: disrupti…
Microsoft released fixes for a record 1,411 CVEs in its October 2026 Patch Tuesday on October 13, more than the roughly 1,100 issued in all of 2025 and above the records set in July and September. The surge is attributed to Microsoft's use of AI-driven vulnerability discovery, including its multi-model agentic scanning harness (MDASH). Th…
Dashlane's third annual Passkey Power 20 report found that passkey use is growing but still trails passwords. The report covers millions of aggregated, anonymized authentications between July 2025 and June 2026. Authentications by Dashlane users nearly doubled year on year to 2.5 million per month, and 57% of users now store at least one …
A coordinated attack on supermarket refrigeration controllers began on the night of 11-12 October 2026 and hit hundreds of stores in France. Attackers remotely forced units into continuous defrost cycles, with temperatures reaching 50°C (122°F), spoiling food and causing fires, according to Paris prosecutors (itsecurityguru.org). The firs…
Researchers at Vedere Labs (Forescout Research) reported on October 14 that they found ROMULUS, a new multi-payload shared ClickFix delivery platform, while investigating an incident. The same malware from a separate ClickFix campaign was linked to another server tied to that incident, and that server hosted the ROMULUS panel. ROMULUS is …
Microsoft, working with industry partners and Japanese police, disrupted a Chinese-language phishing-as-a-service (PhaaS) ecosystem centered on Yam Store, an underground marketplace tracked as Storm-3119. The phishing kits sold there were responsible for over a billion phishing emails and texts sent to more than 75,000 organizations since…
The UK Information Commissioner's Office (ICO) has issued a reprimand and an enforcement notice to the Metropolitan Police Service (MPS) over governance failures in its use of the third-party Auror retail crime intelligence-sharing platform. Between March 2023 and September 2024, MPS officers and staff used Auror without formal approval, …
Check Point researchers disclosed two high-risk vulnerabilities in LiteLLM, an open-source LLM gateway with almost 100,000 downloads a month. The gateway connects enterprise AI applications to providers such as OpenAI, Google and Anthropic. The flaws, CVE-2026-35029 and CVE-2026-59823, could let any user with an account on a vulnerable de…
Censys observed a 71% drop in internet-exposed Danfoss StoreView control panels in France in three days after the October 11 cyber-attack on hundreds of French supermarkets. The attack targeted refrigeration controllers that Danfoss confirmed were mostly older models without security updates (AK-SM800, AK-SC355) or newer models on outdate…
Symantec's Threat Hunter Team reports that new versions of the Aoki infostealer now target credential and session files used by AI coding assistants. Earlier versions focused on browser, cryptocurrency wallet and password manager data. Aoki is a relatively new and sparsely documented stealer, first observed earlier in 2026, and its AI-rel…
A cybersecurity incident at US government contractor BAE Systems has knocked out the systems behind the Bankruptcy Noticing Center (BNC), which generates and distributes bankruptcy notices for all federal bankruptcy courts in the US. The Administrative Office of the US Courts (AOUSC) confirmed on October 15 that BNC IT systems and website…
The Semperis 2026 Ransomware Risk Report, published on October 14, warns that ransomware operators are increasingly threatening the physical safety of CISOs, executives and other senior staff at victim organizations to force payment. The report says a quarter of ransomware campaigns now involve psychological or physical threats against em…
Google warns that Armory Castle, a suspected PRC state-sponsored espionage actor, exploited the critical BeyondTrust Remote Support vulnerability CVE-2026-102764 (CVSS 9.3) in August 2026, well before BeyondTrust disclosed it on October 15. The actor escalated privileges to root and established persistent access on compromised appliances.…
US cybersecurity job postings fell 26.4% year-over-year between January and September 2026, according to a CyberSN analysis of JobSpikr data. The 235,786 roles posted are the lowest total in the four years CyberSN has run the analysis, against 273,498 in 2023, 244,911 in 2024 and 320,210 in 2025. Leadership roles fell sharply: CISO postin…
OpenCode, the open-source AI agent framework developed by Anomaly, confirmed on October 17 a security breach that occurred the night before and affected its Console service. Console stores each workspace's third-party AI provider credentials centrally. About 1.4% of Console users, those with external providers linked, may have had OpenAI,…
A Moody's report published on October 19 finds that known exploited vulnerabilities take a median of 87 days to be patched, leaving half unremediated three months after disclosure. Vulnerabilities with no known exploitation evidence take longer, at a median of 107 days. The report warns that AI-powered vulnerability discovery is driving a…
The Cisco Splunk 2026 State of Observability report, published on October 20, finds that organizations are deploying AI agents faster than they can build monitoring procedures and security controls. This leaves them exposed to security threats, hallucinations and related problems. Of the 2050 ITOps and engineering respondents, 38% said th…
