A short Bloomberg report published in late October 2026 carries a blunt message. According to an executive at Anthropic PBC, mass cybersecurity disruptions will "get a lot worse before they get better." The source material is brief, but the warning is worth taking seriously, and it raises a practical question for ordinary readers: with AI cyberattacks, how to protect yourself without falling into panic or buying the wrong fix?
This post covers what the warning says, what it plausibly means for everyday threats, and which habits still hold up. It also looks at where a VPN fits in, and where it does not.
What the Anthropic Warning Actually Says
The Bloomberg piece, headlined "AI and Cybersecurity: Why Cyberattacks Are Getting Harder to Stop," summarizes a single claim: an Anthropic executive expects mass cybersecurity disruptions to worsen before conditions improve. The excerpt available to us does not include specific incidents, figures, timelines, or technical details, so we will not guess at them.
What the statement does tell us is the direction of travel from someone working inside an AI company. The phrase "before they get better" is also worth noting. It implies a transition period, not a permanent state. That framing suggests defenders are expected to catch up, but not immediately.
For readers, the useful takeaway is not a prediction of doom. It is a reminder to tighten basic defenses now, while the gap between attackers and defenders is being debated.
How AI Is Lowering the Cost of Phishing, Credential Theft, and Malware
The source article does not detail the mechanics, so what follows is general context on how AI tools are widely understood to affect attackers, not claims drawn from the Bloomberg report.
The core idea is cost and scale. Tasks that once required time and skill can be partly automated:
- Phishing: Messages can be written quickly, in many languages, and tailored to a target. The old advice to look for clumsy grammar is less reliable than it used to be.
- Credential theft: Fake login pages and convincing pretexts can be produced in bulk, making stolen passwords easier to harvest across many victims at once.
- Malware: Automation can help less experienced attackers modify or assemble malicious code faster.
The common thread is that a single attacker can attempt more, with less effort. That is why the warning speaks of mass disruption instead of isolated incidents. Our weekly threats roundup on zero-days, AI agents, and data leaks tracks how these themes keep appearing together.
Practical Steps That Still Work: MFA, Password Managers, and Updates
The encouraging part is that the defenses that blunt automated attacks are mostly the familiar ones. Cheap, scaled attacks tend to rely on weak passwords, reused credentials, and unpatched software. Closing those doors removes many easy targets.
Turn on multi-factor authentication (MFA). Even if a convincing fake page captures your password, a second factor adds a barrier. Where possible, prefer app-based or hardware-based methods over text messages.
Use a password manager. A manager generates long, unique passwords for every account, which limits the damage when one site is breached. It can also help you spot fakes: a manager typically will not autofill credentials on a lookalike domain.
Install updates promptly. Operating systems, browsers, and apps receive security patches regularly. Enabling automatic updates shrinks the window in which known flaws can be exploited.
Slow down on unexpected requests. Because AI-written messages can look polished, judge them by context rather than polish. If a message creates urgency, asks for money, or requests a login, verify it through a separate channel, such as typing the official site address yourself or calling a known number.
Remove what you no longer use. Old accounts and unused apps are extra surface area. Deleting them reduces what can be breached or exploited.
Where a VPN Fits in a Layered Defense, and Where It Doesn't
A VPN encrypts traffic between your device and the VPN server and hides your IP address from the sites you visit. That is genuinely useful on public Wi-Fi and for limiting what your internet provider can see.
But it is a limited tool against the threats described above. A VPN does not stop you from entering your password on a phishing page. It does not prevent you from downloading malware, and it does not patch vulnerable software. If an AI-written email convinces you to click, the VPN will faithfully encrypt that connection to the attacker's site.
Think of a VPN as one layer that protects data in transit, not a shield against scams or credential theft. If you use one, treat it as a complement to MFA, a password manager, and regular updates, not a replacement.
What This Means For You
The Anthropic warning does not change what good security looks like, but it raises the stakes of skipping the basics. If attackers can run phishing and credential theft more cheaply, the likeliest victims are people with reused passwords, no second factor, and outdated devices.
The good news is that these are fixable in an afternoon. You do not need to understand the underlying AI to benefit from defenses that make you a harder, less profitable target.
Takeaways: AI Cyberattacks and How to Protect Yourself Today
If you are wondering about AI cyberattacks and how to protect yourself, start with these steps today:
- Enable MFA on your email, banking, and social accounts first.
- Set up a password manager and replace reused passwords, starting with your most important accounts.
- Turn on automatic updates for your phone, computer, and browser.
- Verify unexpected requests through a separate channel before clicking or paying.
- Use a VPN if it suits your needs, but do not rely on it alone.
For more on how AI-driven risks are evolving, check our weekly threats roundup, and revisit your settings regularly as the situation develops.
