As we step into an era where artificial intelligence (AI) plays an increasingly significant role in cybersecurity, discussions surrounding its offensive capabilities are becoming more prominent. A recent report by Anthropic—a leading AI research lab—has sparked the latest conversation on this topic, with questions raised about their claim that an AI-assisted attack they observed was “90% autonomous.”
Critics argue that the report didn’t provide sufficient detail to understand the tools used or the methodologies employed during these attacks. While the report may lack specific detail concerning the AI-assisted attacks, IBM X-Force recognizes the following, perhaps overlooked, important key takeaways: Current cyber capabilities observed today are merely side effects of AI models trained on coding datasets.
Read more…
Source: IBM
Sign up for the Cyber Security Review Newsletter
The latest cyber security news and insights delivered right to your inbox
Related:
- Top AI tools such as OpenClaw and Github Copilot can be hijacked to create new massive botnets
July 15, 2026
Your favorite AI service could be subverted to deploy code that turns your phone or PC into a botnet, according to researchers at Intuit, Technion, and Tel Aviv University. The technique has been given the name HalluSquatting, a portmanteau of adversarial hallucination squatting, and is similar to typosquatting in that it relies on a mistake in ...
- Warning: Scammers are using FaceTime to empty bank accounts
July 14, 2026
Apple is urging users to treat any suspicious FaceTime call or message as untrusted, especially if it involves payments, refunds, password resets, or requests for personal information. This warning appears in a broader Apple support article about scams that target iPhone and iPad users through social engineering. Apple says attackers may contact people by phone calls, FaceTime, text ...
- Microsoft tops last month’s record with 622 Patch Tuesday CVEs
July 14, 2026
Remember last month when we were awed by Microsoft’s record-setting Patch Tuesday that addressed 206 CVEs? That was a quaint era compared to this month: Redmond just rolled out patches for 622 CVEs specific to its products, slightly more than tripling last month’s all-time high. Redmond’s Patch Tuesday release is once again one for the record books, with everything under ...
- Ghostcommit attack hides malicious AI instructions in images
July 13, 2026
Ghostcommit is a proof of concept that shows how AI assistants used to review software code can be tricked by hidden instructions embedded in images. The academic ASSET Research Group showed that an attacker can place instructions inside an image file, point to it in an AGENTS.md file, and get an AI coding agent to follow those instructions during a ...
- EU-UK sanctions target Russia over Europe-wide ‘vast cyber campaign’
July 13, 2026
The European Union and the UK have announced coordinated sanctions against Russia after accusing Moscow’s FSB intelligence agency of carrying out a cyber attack in December targeting Poland’s energy grid and orchestrating a wider campaign of digital sabotage across Europe. The joint measures focus on individuals and organisations linked to Russia’s security services, with the EU ...
- Trump administration subpoenas New York Times journalists over new Air Force One reporting
July 11, 2026
The Trump administration has issued subpoenas to several New York Times journalists after the newspaper reported on security concerns with the president’s new plane. The Times said its journalists were subpoenaed on Friday by the US justice department to testify before a federal grand jury in Manhattan five days later, marking the latest effort by the Trump White ...