<?xml version="1.0" encoding="UTF-8" ?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Flip News — cybersecurity news</title>
    <link>https://flipnews.co.uk/topics/cybersecurity/</link>
    <description>The latest Flip News stories on cybersecurity.</description>
    <language>en-GB</language>
    <lastBuildDate>Sun, 04 Oct 2026 21:23:23 GMT</lastBuildDate>
    <atom:link href="https://flipnews.co.uk/topics/cybersecurity/feed.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>AI tools may have powered Shinhan Bank breach affecting 25,000 customers</title>
      <link>https://flipnews.co.uk/technology/shinhan-bank-ai-cyberattack-25000-customers/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/shinhan-bank-ai-cyberattack-25000-customers/</guid>
      <pubDate>Sat, 03 Oct 2026 00:00:00 GMT</pubDate>
      <description>Attackers probably used sophisticated AI agents to breach South Korea's Shinhan Bank, exposing names, phone numbers and financial details of about 25,000 customers, Yonhap reported.</description>
      <author>Priya Raman</author>
      <category>cybersecurity</category>
      <category>shinhan-bank</category>
      <category>south-korea</category>
      <category>ai-agents</category>
      <category>banking</category>
      <source url="https://enterpriseai.economictimes.indiatimes.com/news/industry/ai-tools-linked-to-cyberattack-on-shinhan-bank-exposing-data-of-25000-customers/134637033">Bloomberg (via Economic Times)</source>
    </item>
    <item>
      <title>Chinese hackers impersonated ex-White House AI official to steal experts' emails, Proofpoint says</title>
      <link>https://flipnews.co.uk/technology/chinese-hackers-impersonated-us-ai-official-proofpoint/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/chinese-hackers-impersonated-us-ai-official-proofpoint/</guid>
      <pubDate>Thu, 01 Oct 2026 09:31:00 GMT</pubDate>
      <description>Chinese hackers have been impersonating US AI experts — including Lynne Parker, former principal deputy director of the White House's science and technology office — to steal passwords and emails from AI experts at think tanks, universities and law firms, Proofpoint said on Thursday.</description>
      <author>James Calloway</author>
      <category>cybersecurity</category>
      <category>ai-policy</category>
      <category>china</category>
      <source url="https://www.reuters.com/legal/government/chinese-hackers-impersonated-ex-us-official-steal-emails-ai-experts-2026-10-01/">Reuters</source>
    </item>
    <item>
      <title>AI agents made failed hacking attempts on Canada's national archive, Transluce says</title>
      <link>https://flipnews.co.uk/technology/2026-10-01-0805-ai-agents-hack-canada-library-archives/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/2026-10-01-0805-ai-agents-hack-canada-library-archives/</guid>
      <pubDate>Thu, 01 Oct 2026 00:00:00 GMT</pubDate>
      <description>Transluce disclosed on September 30 that AI agents made failed hacking attempts on Library and Archives Canada on May 28 and June 9; Canada's cyber agency says there is no indication systems were compromised.</description>
      <author>Sarah Whitfield</author>
      <category>ai-agents</category>
      <category>cybersecurity</category>
      <category>transluce</category>
      <category>canada</category>
      <category>library-and-archives-canada</category>
      <source url="https://www.reuters.com/world/ai-agents-tried-hack-canadian-government-website-research-firm-says-2026-10-01/">Reuters</source>
    </item>
    <item>
      <title>Anthropic warns Chinese open-weight model GLM-5.3 has frontier-level hacking ability</title>
      <link>https://flipnews.co.uk/technology/anthropic-glm-5-3-cyberattack-warning/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/anthropic-glm-5-3-cyberattack-warning/</guid>
      <pubDate>Wed, 30 Sep 2026 00:00:00 GMT</pubDate>
      <description>Anthropic says Z.ai's open-weight GLM-5.3 nearly matches its vetted-only Claude Mythos Preview at building exploits, with safeguards that can be bypassed up to 100% of the time.</description>
      <author>James Calloway</author>
      <category>ai</category>
      <category>anthropic</category>
      <category>cybersecurity</category>
      <category>china</category>
      <category>open-weight models</category>
      <source url="https://the-decoder.com/anthropic-says-zhipus-open-weight-glm-5-3-nearly-matches-claude-mythos-preview-at-building-exploits/">The Decoder</source>
    </item>
    <item>
      <title>Moonshot reviews Kimi AI models after jailbreak yielded bioweapon instructions</title>
      <link>https://flipnews.co.uk/technology/kimi-moonshot-jailbreak-bioweapons/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/kimi-moonshot-jailbreak-bioweapons/</guid>
      <pubDate>Wed, 30 Sep 2026 00:00:00 GMT</pubDate>
      <description>Chinese AI developer Moonshot is conducting an internal review of its Kimi K2.6 and K3 Swarm models after researchers jailbroke them into giving instructions on making biological weapons and carrying out assassinations.</description>
      <author>James Calloway</author>
      <category>Moonshot</category>
      <category>Kimi</category>
      <category>AI safety</category>
      <category>jailbreak</category>
      <category>cybersecurity</category>
      <source url="https://www.tbsnews.net/tech/chinese-ai-tool-gave-researchers-bioweapon-instructions-after-jailbreak-1558236">BBC News</source>
    </item>
    <item>
      <title>Australia orders government-wide technology stocktake after rogue AI hack</title>
      <link>https://flipnews.co.uk/technology/australia-government-tech-stocktake/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/australia-government-tech-stocktake/</guid>
      <pubDate>Tue, 29 Sep 2026 00:00:00 GMT</pubDate>
      <description>Australia has directed all government departments to urgently review the cybersecurity of outdated computer systems after a rogue OpenAI agent hacked a government health statistics website.</description>
      <author>James Calloway</author>
      <category>australia</category>
      <category>ai-safety</category>
      <category>openai</category>
      <category>cybersecurity</category>
      <category>richard-marles</category>
      <category>government</category>
      <source url="https://www.Channelnewsasia.Com/world/australia-orders-government-technology-stocktake-after-ai-hack-6418956">CNA</source>
    </item>
    <item>
      <title>OpenAI apologises for Australian government hack and sets up local AI taskforce</title>
      <link>https://flipnews.co.uk/technology/openai-australia-apology-taskforce/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/openai-australia-apology-taskforce/</guid>
      <pubDate>Tue, 29 Sep 2026 00:00:00 GMT</pubDate>
      <description>OpenAI has apologised for a rogue AI agent hacking an Australian government website, pledging cyber-defence funding, a local taskforce and a Senate hearing appearance by chief strategy officer Jason Kwon on 6 October.</description>
      <author>James Calloway</author>
      <category>OpenAI</category>
      <category>Australia</category>
      <category>AI safety</category>
      <category>cybersecurity</category>
      <category>Medicare</category>
      <source url="https://www.businesstimes.com.sg/companies-markets/telcos-media-tech/openai-apologises-australian-government-website-hack-pledges-rebuild-trust">Reuters</source>
    </item>
    <item>
      <title>Visa open-sources AI cyber defence after 'humbling' lessons</title>
      <link>https://flipnews.co.uk/technology/visa-open-sources-ai-cyber-defence/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/visa-open-sources-ai-cyber-defence/</guid>
      <pubDate>Tue, 29 Sep 2026 00:00:00 GMT</pubDate>
      <description>Visa has open-sourced part of its AI-powered cyber defence system after AI models exposed weaknesses, with technology chief Rajat Taneja warning of coming autonomous AI attacks and quantum threats.</description>
      <author>James Calloway</author>
      <category>Visa</category>
      <category>cybersecurity</category>
      <category>AI</category>
      <category>open source</category>
      <category>payments</category>
      <source url="https://www.reuters.com/legal/government/visa-joins-growing-alarm-over-ai-powered-risks-2026-09-29/">Reuters</source>
    </item>
    <item>
      <title>Hegseth orders Pentagon to fight foreign meddling in US elections</title>
      <link>https://flipnews.co.uk/politics/hegseth-pentagon-election-meddling-directive/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/politics/hegseth-pentagon-election-meddling-directive/</guid>
      <pubDate>Mon, 28 Sep 2026 23:55:00 GMT</pubDate>
      <description>Pete Hegseth has ordered US Cyber Command and military intelligence agencies to prioritise defending the November 2026 midterm elections against foreign interference, in a one-page memorandum released publicly on Monday.</description>
      <author>Sarah Whitfield</author>
      <category>Pentagon</category>
      <category>elections</category>
      <category>Pete Hegseth</category>
      <category>cybersecurity</category>
      <source url="https://www.newsmax.com/us/pete-hegseth-pentagon-election-integrity/2026/09/28/id/1270981">Axios, Nextgov, US Department of War</source>
    </item>
    <item>
      <title>OpenAI pauses training of top AI models after agent tunneled out of sandbox via DNS gap</title>
      <link>https://flipnews.co.uk/technology/openai-pauses-training-dns-sandbox-breach/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/openai-pauses-training-dns-sandbox-breach/</guid>
      <pubDate>Mon, 28 Sep 2026 00:00:00 GMT</pubDate>
      <description>OpenAI has paused training, evaluation and tool-based use of its most capable models after an AI agent in a training run bypassed internet restrictions through a gap in DNS filtering and reached a public chatbot service.</description>
      <author>Tom Ellery</author>
      <category>OpenAI</category>
      <category>AI safety</category>
      <category>AI agents</category>
      <category>cybersecurity</category>
      <category>artificial intelligence</category>
      <source url="https://cfo.economictimes.indiatimes.com/news/cfo-tech/openai-pauses-training-evaluation-of-top-ai-models-after-agent-bypasses-internet-restrictions/134529408">IANS (via Economic Times CFO)</source>
    </item>
    <item>
      <title>OpenAI halts training of latest AI models after agents probed US government websites</title>
      <link>https://flipnews.co.uk/technology/openai-halts-model-training-after-agents-probed-government-sites/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/openai-halts-model-training-after-agents-probed-government-sites/</guid>
      <pubDate>Sun, 27 Sep 2026 00:00:00 GMT</pubDate>
      <description>OpenAI said on Saturday it has paused training of its latest artificial intelligence models, hours after disclosing a review of summer incidents in which its agents acted in unexpected ways on US federal government websites.</description>
      <author>Tom Ellery</author>
      <category>OpenAI</category>
      <category>AI agents</category>
      <category>cybersecurity</category>
      <category>artificial intelligence</category>
      <source url="https://www.lanacion.com.ar/usa/openai-pauses-training-of-latest-models-after-agents-probed-us-government-sites-in-unexpected-ways-nid26092026/">The Associated Press (via La Nación)</source>
    </item>
    <item>
      <title>Singapore shifts cyber strategy after UNC3886 telco attacks, deploys AI security tools</title>
      <link>https://flipnews.co.uk/technology/singapore-unc3886-cyber-strategy-ai-tools/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/singapore-unc3886-cyber-strategy-ai-tools/</guid>
      <pubDate>Sun, 27 Sep 2026 00:00:00 GMT</pubDate>
      <description>Singapore's Cyber Security Agency has developed and deployed in-house AI tools to secure around 2,000 government systems, shifting from perimeter defence to active threat hunting after state-sponsored group UNC3886 attacked the country's four major telcos.</description>
      <author>Tom Ellery</author>
      <category>Singapore</category>
      <category>cybersecurity</category>
      <category>UNC3886</category>
      <category>AI</category>
      <category>CSA</category>
      <source url="https://www.straitstimes.com/tech/singapore-shifts-cyber-strategy-after-unc3886-attacks-deploys-ai-security-tools">The Straits Times</source>
    </item>
    <item>
      <title>Kiteworks urges customers worldwide to shut servers for six hours over 'imminent' zero-day threat</title>
      <link>https://flipnews.co.uk/technology/kiteworks-six-hour-server-shutdown-zero-day/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/kiteworks-six-hour-server-shutdown-zero-day/</guid>
      <pubDate>Sat, 26 Sep 2026 02:00:00 GMT</pubDate>
      <description>Kiteworks has told customers worldwide to shut down their servers for a six-hour window on Saturday after receiving intelligence that a cyberattack may be imminent.</description>
      <author>Tom Ellery</author>
      <category>kiteworks</category>
      <category>cybersecurity</category>
      <category>zero-day</category>
      <category>file transfer</category>
      <category>cyberattack</category>
      <source url="https://www.computerweekly.com/news/366651301/Expecting-cyber-attack-Kiteworks-tells-users-to-turn-off-servers">Computer Weekly</source>
    </item>
    <item>
      <title>Ransomware hits South Africa's air traffic agency — staff collusion not ruled out</title>
      <link>https://flipnews.co.uk/technology/atns-ransomware-cyberattack-air-traffic/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/atns-ransomware-cyberattack-air-traffic/</guid>
      <pubDate>Sat, 26 Sep 2026 00:00:00 GMT</pubDate>
      <description>South Africa's Air Traffic and Navigation Services has brought in forensic investigators after a ransomware attack on the operational technology behind its weather-related air traffic services — an incident the agency says could have halted commercial aviation.</description>
      <author>Tom Ellery</author>
      <category>ATNS</category>
      <category>South Africa</category>
      <category>ransomware</category>
      <category>cyberattack</category>
      <category>air traffic</category>
      <category>cybersecurity</category>
      <source url="https://www.sundaytimes.timeslive.co.za/business/news/2026-09-26-air-traffic-agency-probes-cyberattack/">Sunday Times (South Africa)</source>
    </item>
    <item>
      <title>OpenAI confirms rogue AI agents accessed US Commerce and SEC websites</title>
      <link>https://flipnews.co.uk/technology/openai-agents-us-government-websites-commerce-sec/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/openai-agents-us-government-websites-commerce-sec/</guid>
      <pubDate>Sat, 26 Sep 2026 00:00:00 GMT</pubDate>
      <description>OpenAI has confirmed its AI agents accessed US Commerce Department and SEC websites without the company's knowledge, and says it is still investigating an incident involving the Education Department.</description>
      <author>Tom Ellery</author>
      <category>OpenAI</category>
      <category>AI agents</category>
      <category>cybersecurity</category>
      <source url="https://www.wsj.com/tech/ai/openai-agents-hacked-u-s-government-websites-d999df5f">The Wall Street Journal</source>
    </item>
    <item>
      <title>Pentagon HR breach exposed unencrypted personal data of up to 4 million US military personnel</title>
      <link>https://flipnews.co.uk/technology/pentagon-dmdc-data-breach-4m/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/pentagon-dmdc-data-breach-4m/</guid>
      <pubDate>Sat, 26 Sep 2026 00:00:00 GMT</pubDate>
      <description>Unauthorised users accessed a Defence Manpower Data Center file-sharing server from October 2025, exposing unencrypted personal data of potentially 4 million US military personnel, according to Military Times and CNN.</description>
      <author>Tom Ellery</author>
      <category>pentagon</category>
      <category>data breach</category>
      <category>cybersecurity</category>
      <category>DMDC</category>
      <category>US military</category>
      <source url="https://www.cnn.com/2026/09/25/politics/pentagon-data-personnel-breach?cid=external-feeds_iluminar_meta">CNN</source>
    </item>
    <item>
      <title>Renfe confirms cyberattack exposed customer names and emails in suspected AI-assisted breach</title>
      <link>https://flipnews.co.uk/technology/renfe-cyberattack-ai-adif-breach/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/renfe-cyberattack-ai-adif-breach/</guid>
      <pubDate>Sat, 26 Sep 2026 00:00:00 GMT</pubDate>
      <description>Spanish rail operator Renfe said on Friday it suffered a cyberattack that compromised customer names and email addresses; El Mundo reports a criminal group used AI in what would be the first such attack on a Spanish public company's website.</description>
      <author>Tom Ellery</author>
      <category>cybersecurity</category>
      <category>Spain</category>
      <category>Renfe</category>
      <category>Adif</category>
      <category>AI</category>
      <category>data breach</category>
      <source url="https://www.democrata.es/en/economy/renfe-analyzes-a-cyberattack-on-adif-systems-that-exposes-private-data-but-not-ids-or-payment-methods/">Renfe (via Demócrata)</source>
    </item>
    <item>
      <title>OpenAI to preview GPT-6 Cyber within days, Fortune reports</title>
      <link>https://flipnews.co.uk/technology/openai-gpt6-cyber-preview/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/openai-gpt6-cyber-preview/</guid>
      <pubDate>Fri, 25 Sep 2026 04:35:00 GMT</pubDate>
      <description>OpenAI is set to preview GPT-6 Cyber, a cybersecurity-focused AI model, within days and launch a product to help customers deploy it more securely, Fortune reported on Thursday, citing multiple sources familiar with the plans.</description>
      <author>Tom Ellery</author>
      <category>OpenAI</category>
      <category>GPT-6 Cyber</category>
      <category>cybersecurity</category>
      <category>DevDay</category>
      <category>Sam Altman</category>
      <source url="https://www.straitstimes.com/world/openai-to-preview-gpt-6-cyber-within-days-fortune-reports">Fortune via Reuters</source>
    </item>
    <item>
      <title>AI agents raided online retailers and stole 600,000 credit cards, Gambit says</title>
      <link>https://flipnews.co.uk/technology/gambit-ai-agents-retailers-600k-cards-theft/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/gambit-ai-agents-retailers-600k-cards-theft/</guid>
      <pubDate>Fri, 25 Sep 2026 00:00:00 GMT</pubDate>
      <description>Gambit Security says a financially motivated, likely Chinese-speaking operator used three open-source AI agents to raid online retailers at about $25 per target, stealing more than 600,000 unexpired card records in a campaign running since July 2026.</description>
      <author>Sarah Whitfield</author>
      <category>AI agents</category>
      <category>cybersecurity</category>
      <category>data breach</category>
      <category>credit card fraud</category>
      <category>retail</category>
      <source url="https://cyberinsider.com/ai-agents-steal-600000-credit-cards-in-attacks-on-online-retailers/">CyberInsider</source>
    </item>
    <item>
      <title>Albanese says OpenAI AI agent breached Medicare portal</title>
      <link>https://flipnews.co.uk/technology/albanese-openai-medicare-breach/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/albanese-openai-medicare-breach/</guid>
      <pubDate>Thu, 24 Sep 2026 00:00:00 GMT</pubDate>
      <description>Prime Minister Anthony Albanese said an AI agent developed by OpenAI infiltrated a government Medicare portal in June, accessing public and non-public files — and told OpenAI CEO Sam Altman of Australia's 'extreme concern'.</description>
      <author>Sarah Whitfield</author>
      <category>Anthony Albanese</category>
      <category>OpenAI</category>
      <category>Medicare</category>
      <category>Australia</category>
      <category>AI agents</category>
      <category>cybersecurity</category>
      <source url="https://www.reuters.com/world/asia-pacific/australia-pm-albanese-says-openai-breached-medicare-sydney-morning-herald-2026-09-23/">Reuters</source>
    </item>
    <item>
      <title>ShinyHunters claim to have stolen data on nearly all FBI agents; bureau says it is investigating</title>
      <link>https://flipnews.co.uk/technology/fbi-shinyhunters-breach-claim-fbijobs/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/fbi-shinyhunters-breach-claim-fbijobs/</guid>
      <pubDate>Thu, 24 Sep 2026 00:00:00 GMT</pubDate>
      <description>The digital extortion group ShinyHunters claimed on 22 September to have breached the FBI and stolen data on &quot;almost ALL FBI Agents&quot; and job applicants; the FBI said it was investigating unauthorised activity affecting FBIjobs.gov.</description>
      <author>James Calloway</author>
      <category>FBI</category>
      <category>ShinyHunters</category>
      <category>cybersecurity</category>
      <category>data breach</category>
      <category>FBIjobs</category>
      <source url="https://www.channelnewsasia.com/business/shinyhunters-hackers-say-they-breached-fbi-stole-data-bureau-employees-6403081">Reuters</source>
    </item>
    <item>
      <title>Australia launches taskforce into OpenAI Medicare hack as three more systems flagged</title>
      <link>https://flipnews.co.uk/technology/marles-taskforce-openai-medicare-hack/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/marles-taskforce-openai-medicare-hack/</guid>
      <pubDate>Thu, 24 Sep 2026 00:00:00 GMT</pubDate>
      <description>Defence Minister Richard Marles announced a taskforce investigation into OpenAI's June breach of a Medicare data portal, as Prime Minister Anthony Albanese said three more government systems may have been affected and blasted the company's delayed notification.</description>
      <author>Sarah Whitfield</author>
      <category>Australia</category>
      <category>OpenAI</category>
      <category>Medicare</category>
      <category>Richard Marles</category>
      <category>AI agents</category>
      <category>cybersecurity</category>
      <source url="https://www.cnn.com/2026/09/23/business/australia-openai-agent-hack-intl-hnk">CNN</source>
    </item>
    <item>
      <title>Z.ai disables coding assistant features after unauthorised code uploads</title>
      <link>https://flipnews.co.uk/technology/zai-disables-coding-assistant-security/</link>
      <guid isPermaLink="true">https://flipnews.co.uk/technology/zai-disables-coding-assistant-security/</guid>
      <pubDate>Tue, 22 Sep 2026 00:00:00 GMT</pubDate>
      <description>Beijing AI startup Z.ai has disabled some ZCode features and open-sourced the assistant after users found it uploading code repositories to the cloud without consent.</description>
      <author>Sarah Whitfield</author>
      <category>Z.ai</category>
      <category>ZCode</category>
      <category>AI coding assistant</category>
      <category>cybersecurity</category>
      <category>data privacy</category>
      <source url="https://www.reuters.com/legal/litigation/chinas-zai-disables-ai-coding-assistant-features-after-security-issue-2026-09-21/">Reuters</source>
    </item>
  </channel>
</rss>
