While retail investors debate how to get exposure to Anthropic ahead of its reported IPO, institutional money has been positioning for months through channels largely unavailable to individual investors. Understanding how pension funds, sovereign wealth vehicles, and specialized pre-IPO platforms are approaching the deal offers a useful blueprint — even if most retail investors can’t fully replicate the strategy.

Key Takeaways

• Anthropic’s last private round — a $65 billion Series H at a $965 billion valuation in May 2026 — was led by Altimeter Capital, Dragoneer, Greenoaks, and other growth-focused institutional investors.
• Existing shareholders face a lockup reportedly running through December 2026, meaning even institutional holders can’t freely sell immediately after listing.
• Institutional investors are reportedly using a two-year forward revenue framework (2028 projections) rather than trailing metrics to justify entry valuations near $2 trillion.
• Secondary market transactions — where existing shareholders or employees sell stakes to new investors before an IPO — have been a key channel for institutional and accredited investor access.
• Free float at listing is expected to be unusually low, meaning institutional positioning before the IPO carries outsized influence over available shares.

Why Institutional Investors Move Earlier — and Differently

Retail investors typically only gain access to a company once it lists publicly, or in rare cases through a limited retail tranche of the IPO itself. Institutional investors, by contrast, have multiple additional entry points that predate the public listing entirely:

1. Primary funding rounds — direct participation in venture and growth-equity rounds, such as Anthropic’s May 2026 Series H
2. Secondary market purchases — buying existing shares directly from early employees, founders, or earlier-round investors seeking liquidity before a lockup
3. Structured pre-IPO funds — pooled vehicles that acquire blocks of private company shares and offer accredited investors indirect exposure
4. Anchor investor allocations — negotiated commitments to purchase a defined block of shares at IPO pricing, arranged directly with the underwriting banks

Inside Anthropic’s Most Recent Institutional Round

Anthropic’s May 28, 2026 Series H round — which raised $65 billion at a $965 billion post-money valuation, more than double its $380 billion valuation in February — was led by a group of growth-stage investors including Altimeter Capital, Dragoneer, and Greenoaks, names well known for late-stage pre-IPO positioning in high-growth technology companies.

This round is instructive for retail investors trying to understand institutional logic: these firms priced their entry at less than half of what bankers are now reportedly discussing for the IPO itself just months later. That’s either validation of extraordinary execution, or a sign of how quickly sentiment (and pricing) can shift in a hot AI cycle — likely some of both.

The Two-Year Forward Framework Institutions Are Using

One of the more unusual aspects of institutional positioning around Anthropic is the valuation framework itself. Rather than the standard “next twelve months” (NTM) forward multiple most public equity investors use, bankers and institutional backers are reportedly using a two-year forward horizon, anchored to 2028 revenue projections of $190–200 billion.

This matters strategically because:

• A one-year forward multiple on Anthropic’s current run rate looks aggressive (~17–20x projected 2026 revenue)
• A two-year forward multiple looks comparatively reasonable (~10x projected 2028 revenue), in line with or cheaper than Nvidia’s current multiple
• Institutions willing to underwrite the longer growth runway can justify materially higher entry prices than those anchored to trailing or near-term metrics

For retail investors evaluating the eventual public stock, understanding which framework the market is using at any given moment — trailing, one-year forward, or two-year forward — is essential to interpreting whether the stock looks “cheap” or “expensive” relative to institutional benchmarks.

Secondary Markets: The Institutional Workaround for Lockups

With existing Anthropic shareholders reportedly locked up through December 2026, institutional investors seeking exposure before then have increasingly turned to structured secondary transactions — privately negotiated purchases of existing shares from early employees or earlier investors, often facilitated by specialized broker-dealers or platforms.

What Retail-Accessible Pre-IPO Platforms Actually Offer

A subset of institutional-style access has become available to accredited (and in limited cases, non-accredited) individual investors through pre-IPO investing platforms. These platforms typically structure exposure through special purpose vehicles (SPVs) or forward purchase contracts rather than direct share ownership, and they come with meaningfully different risk characteristics than buying stock on the open market:

• Higher fees — placement fees and carried interest that reduce net returns relative to direct share ownership
• Illiquidity — positions often can’t be sold until the underlying company lists or a secondary window opens
• Valuation opacity — SPV pricing may not perfectly track the company’s actual last-round valuation
• Accreditation requirements — many platforms restrict access to investors meeting SEC accredited investor income or net worth thresholds

How Institutional Positioning Could Affect the IPO Itself

The scale of institutional demand ahead of the offering has a direct mechanical effect on how the deal gets priced. If Morgan Stanley and Goldman Sachs’s bookbuilding process shows overwhelming institutional demand at or above the reported $2 trillion target, it strengthens the case for pricing at or near the top of any eventual range. Conversely, if institutional appetite proves more measured once real due diligence begins on audited (rather than investor-relayed) financials, it could pressure the final offer price downward from current speculative levels.

Lessons Retail Investors Can Actually Apply

While most individual investors can’t access Series H-style rounds or secondary share purchases, a few institutional principles translate directly:

1. Think in multi-year revenue terms, not just trailing metrics, when evaluating whether a post-IPO valuation looks reasonable.
2. Understand the lockup calendar. A December 2026 lockup expiry means a wave of newly tradable shares could hit the market months after listing — a potential source of added volatility worth tracking even for investors who buy on the open market.
3. Don’t mistake institutional participation for a valuation guarantee. Even sophisticated growth investors who led the Series H priced their entry at less than half of the currently discussed IPO target — a reminder that institutional money is not infallible on pricing.

FAQ

Who led Anthropic’s most recent private funding round?

Altimeter Capital, Dragoneer, and Greenoaks led Anthropic’s $65 billion Series H round in May 2026, which valued the company at $965 billion.

Can retail investors access pre-IPO shares the same way institutions do?

Not directly in most cases. Primary funding rounds and secondary share purchases are typically restricted to institutional and accredited investors, though some pre-IPO platforms offer indirect, fee-bearing exposure to accredited individual investors.

Why does the lockup period matter for investors?

A lockup restricts existing shareholders from selling shares for a defined period after an IPO. Anthropic’s lockup is reportedly set to run through December 2026, meaning a significant supply of shares could become tradable months after the initial listing, potentially affecting the stock price.

What valuation framework are institutions using to justify $2 trillion?

Reporting indicates bankers and institutional investors are using a two-year forward revenue projection (targeting 2028 revenue of $190–200 billion) rather than a standard one-year forward multiple, which makes the headline valuation look more justified on a longer time horizon.

The FBI disrupted a Chinese state-sponsored hacking operation that breached NASA, the Senate, and the Federal Reserve. Here’s what happened, why it matters for private-sector cybersecurity, and how to protect your organization.

Key Takeaways

• The FBI and DOJ disrupted a Chinese state-sponsored hacking operation on August 27, 2026, seizing two platforms — “QScan” and “QTRouter” — used to breach NASA, the U.S. Senate, the Federal Reserve, the Department of Justice, and other critical networks.
• The campaign dates back to at least 2018, representing sustained, long-term espionage infrastructure rather than a single breach event.
• Confirmed victims span finance, legislative, scientific, and healthcare sectors, including Department of Energy national laboratories, the National Institutes of Health, hospitals, telecommunications providers, and power utilities.
• The obfuscation technique is particularly notable: QTRouter allowed attackers to route traffic through already-compromised devices, making attacks appear to originate from nearby or domestic sources rather than overseas.
• No individual indictments accompanied the announcement — officials characterized the action as a disruption operation, not a completed prosecution, meaning the underlying threat actors remain at large.

What Happened: The QScan and QTRouter Takedown

On August 27, 2026, the FBI, in coordination with the Department of Justice, announced it had disrupted a long-running, China-affiliated hacking operation by seizing two pieces of malicious infrastructure:

• QScan — a vulnerability scanning and exploitation malware tool used to identify weaknesses in target networks
• QTRouter — an obfuscation network that routed attack traffic through compromised third-party devices, disguising the true origin of intrusions

According to a joint cybersecurity advisory from the FBI, NSA, and U.S. Cyber Command’s Cyber National Mission Force, the operators behind this infrastructure — tracked under the identifier QTFY — conducted a sustained campaign of intrusions and reconnaissance dating back to at least 2018.

A Timeline of Confirmed Activity

• August 2019: An unsuccessful attempt to breach NASA’s servers by exploiting a VPN vulnerability.
• May 2024: Confirmed data theft from defense contractors, financial institutions, and universities.
• September 2024: Successful intrusions into three Department of Energy national laboratories, the National Institutes of Health, an HHS agency component, and a U.S. security-device manufacturer.
• March 2026: Unsuccessful vulnerability scans targeting the U.S. Senate and an American hospital system.
• June 2026: A vulnerability scan of an unidentified U.S. election system.

Why the Obfuscation Technique Matters

Cybersecurity experts have flagged QTRouter’s routing technique as particularly significant. As one cybersecurity company vice president explained, when an intrusion appears to come from a device physically near the target — rather than from overseas — it buys the attacker time and makes attribution significantly slower. This technique effectively weaponized already-compromised consumer and business devices as unwitting relay points, complicating incident response for defenders across multiple victim organizations simultaneously.

Why This Matters Beyond Government Networks

While headlines have focused on high-profile targets like NASA, the Senate, and the Federal Reserve, the practical lesson for the private sector is more expansive. The same campaign also compromised:

• Hospitals and healthcare networks
• Telecommunications providers
• Power utilities
• Universities
• Defense contractors and financial institutions

This breadth illustrates a critical point for private-sector risk managers: nation-state hacking infrastructure does not distinguish neatly between government and private targets. The same tools, techniques, and obfuscation infrastructure used against a federal agency can just as easily be deployed against a mid-sized healthcare system, a regional utility, or a private financial services firm — and in this case, was.

The “Disruption, Not Prosecution” Distinction

Officials explicitly characterized this action as a disruption operation rather than a completed prosecution — no individual indictments were announced alongside the domain seizures. This is an important distinction for organizations assessing ongoing risk: the underlying threat actors and their broader capabilities have not been eliminated, only this specific piece of enabling infrastructure has been degraded. Historical precedent with similar state-sponsored groups suggests operators frequently rebuild alternative infrastructure following takedowns of this kind.

What This Means for Private-Sector Cybersecurity Strategy

1. Assume Nation-State Techniques Will Trickle Down

Techniques pioneered by well-resourced, state-sponsored actors — such as QTRouter’s device-relay obfuscation — often become templates that less sophisticated criminal groups eventually adopt or purchase access to. Organizations should not assume that “we’re not a government target” provides meaningful protection.

2. Device-Level Compromise Is a Systemic Risk

Because QTRouter relied on routing traffic through already-compromised devices — potentially including consumer routers, IoT devices, or under-secured business network equipment — any internet-connected device with weak security hygiene can become part of an attack against an unrelated third party. This underscores the importance of:

• Regular firmware and security patching for all network-connected devices
• Network segmentation to limit lateral movement if any single device is compromised
• Monitoring for unusual outbound traffic patterns that could indicate a device is being used as a relay point

3. Sector-Specific Exposure Requires Sector-Specific Preparedness

Given that hospitals, utilities, telecommunications providers, and financial institutions were all confirmed victims in this specific campcampaign, organizations in these sectors should treat nation-state-level threat modeling as a baseline requirement, not an aspirational upgrade.

Actionable Cybersecurity Takeaways for Organizations

• Review and patch VPN infrastructure immediately. The original 2019 NASA intrusion attempt exploited a VPN vulnerability — a category of exposure that remains a common entry point for state-sponsored actors.
• Implement network traffic anomaly detection capable of identifying unusual routing patterns, particularly traffic that may indicate a device is being used to relay attacks against third parties.
• Conduct third-party and vendor risk assessments with particular attention to any connected devices or systems that might be leveraged as intermediate infrastructure in a broader attack chain.
• Maintain updated cyber insurance coverage that accounts for nation-state-level threat scenarios, given the demonstrated breadth of sectors targeted in this campaign.
• Develop and regularly test incident response plans that account for the possibility of long-dwelling, difficult-to-attribute intrusions, given this campaign’s multi-year operational history before detection and disruption.
• Monitor CISA, FBI, and NSA joint cybersecurity advisories directly, as these often contain specific indicators of compromise (IOCs) that can be used to scan internal networks for related activity.

Frequently Asked Questions

What is the QScan and QTRouter hacking operation? QScan and QTRouter were two hacking platforms used by a China-affiliated threat actor group to scan for vulnerabilities and obfuscate the origin of cyberattacks against U.S. government and critical infrastructure targets, including NASA, the U.S. Senate, and the Federal Reserve, dating back to at least 2018; the FBI and DOJ seized the underlying domains in August 2026.

Did the hackers steal data from NASA? Reporting indicates an attempt to breach NASA’s servers by exploiting a VPN vulnerability in August 2019 was unsuccessful; the broader campaign did successfully compromise other targets, including Department of Energy national laboratories, the National Institutes of Health, and various hospitals, telecommunications providers, and financial institutions over its multi-year operation.

How can my organization protect itself from similar nation-state cyberattacks? Cybersecurity experts recommend patching VPN and network infrastructure regularly, implementing traffic anomaly detection to identify devices potentially being used as attack relays, conducting third-party risk assessments, and maintaining an incident response plan built around long-dwelling, difficult-to-attribute threats rather than assuming only high-profile organizations are targeted.

X’s latest round of privacy policy and data-handling updates has reignited a familiar question for millions of users: how much of your activity on the platform is actually private, and what can you realistically do about it? Between expanded data usage for AI model training, updated location and ad-targeting permissions, and changes to direct message encryption, the platform’s privacy posture in 2026 looks meaningfully different than it did even two years ago. For users who care about limiting exposure — journalists, businesses, or just privacy-conscious individuals — understanding these changes and pairing them with the right tools, including a quality VPN, has become more important than ever.

This guide breaks down what actually changed in X’s privacy settings this year, what data is being collected and how it’s used, and which VPN services offer the best real-world protection for social media privacy in 2026. If you’ve been putting off a proper privacy audit of your social accounts, this is the moment to do it.Privacy

What Changed in X’s 2026 Privacy Policy

The most significant shift has been around AI training data usage — X has expanded how user posts, interactions, and in some cases direct messages can be used to train its AI systems, with opt-out mechanisms that are available but not always prominently surfaced in account settings. Location data handling has also been updated, with more granular ad-targeting permissions tied to device-level location history rather than just IP-based approximation. For users who haven’t reviewed their privacy settings recently, several previously off default toggles have shifted to opt-in by default with updates, which is a common pattern across social platforms and one worth checking manually rather than assuming your old preferences carried over.

It’s worth noting that policy language and actual enforcement don’t always move in lockstep — several privacy researchers have flagged gaps between what X’s policy states and what’s technically observable in data flows, which is part of why relying solely on in-app settings isn’t a complete privacy strategy.

Why a VPN Still Matters, Even With Platform-Level Privacy Settings

A VPN doesn’t change what X itself collects once you’re logged in and posting — that’s governed entirely by the platform’s own data policy. What a VPN does protect is everything happening around your X usage: your real IP address and approximate location being visible to the platform and to your internet service provider, your traffic being visible on public Wi-Fi networks, and third-party trackers embedded across the web being able to correlate your browsing activity outside of X with your identity there.

What a VPN Actually Protects Against

• IP-based location tracking – Masks your real IP address so location isn’t inferred from your connection
• ISP-level monitoring – Prevents your internet provider from logging which sites and platforms you access
• Public Wi-Fi vulnerabilities – Encrypts your traffic on unsecured networks like cafes, airports, and hotels
• Cross-site tracking correlation – Makes it harder for advertisers to link your activity across multiple platforms
• Regional content and access restrictions – Allows access to X in regions where it may face throttling or restrictions

Best VPNs for Social Media Privacy in 2026

Not all VPNs are built equally, and for social media privacy specifically, you want a provider with a verified no-logs policy, strong encryption standards, and fast enough speeds to not disrupt real-time browsing and video content.

Top VPN Picks for 2026

• ProtonVPN – Strong privacy-first reputation, based in Switzerland’s strict data protection jurisdiction, solid free tier
• Mullvad – No email or personal information required to sign up, anonymous account number system, excellent for maximum anonymity
• NordVPN – Best balance of speed and privacy features, independently audited no-logs policy, large server network
• ExpressVPN – Consistently fast speeds, strong for streaming and social media use without lag, audited security practices
• Surfshark – Budget-friendly with unlimited simultaneous device connections, solid for households or small teams

VPN Comparison Table

A Quick Privacy Checklist for X Users

• Review your data-sharing and AI training opt-out settings directly in X’s privacy dashboard, not just the initial prompt
• Turn off precise location sharing unless it’s actively needed for a specific feature
• Use a VPN consistently, not just occasionally, since intermittent use still exposes your real IP most of the time
• Enable two-factor authentication using an authenticator app rather than SMS, which is more vulnerable to interception
• Periodically review connected third-party apps with access to your X account and revoke anything unused

Mobile vs Desktop Privacy Considerations

Privacy exposure isn’t identical across devices, and it’s worth treating your mobile X usage as a separate consideration from desktop browsing. Mobile apps often request additional permissions — precise location, contact list access, camera and microphone — that a browser-based session simply doesn’t have access to, meaning your phone’s app-level permissions matter as much as any VPN or in-app privacy setting. Most reputable VPN providers now offer dedicated mobile apps that encrypt your device’s traffic system-wide rather than just within a single browser, which is important since a browser-only VPN extension won’t protect traffic from the native X app. Reviewing your phone’s app permission settings for X directly, alongside your VPN and in-platform privacy settings, closes a gap that many privacy-conscious users overlook by focusing exclusively on browser-based protections.

Frequently Asked Questions

Does a VPN make me completely anonymous on X?

No — a VPN protects your IP address and network-level traffic, but you’re still identifiable through your account login, posting patterns, and any personal information in your profile or content. True anonymity requires a combination of measures well beyond just a VPN, including careful account hygiene and avoiding identifying details in your posts.

Can X detect that I’m using a VPN?

Some platforms can detect VPN usage through IP reputation databases, and in rare cases this can trigger additional verification steps or regional content restrictions. Reputable VPN providers with large, frequently rotated server networks generally minimize this friction better than smaller or free VPN services.

Is a free VPN good enough for social media privacy?

Generally not recommended for anything beyond casual use. Free VPNs often monetize through data logging or ad injection, which directly undermines the privacy goal you’re trying to achieve, and their server networks and speeds are typically far more limited than paid, audited providers.

Do I need a VPN if I’ve already adjusted all my X privacy settings?

Yes, because privacy settings and a VPN protect different things. In-app settings control what X itself does with your data and how visible your content is to other users, while a VPN protects your network-level identity and traffic from your ISP and other third parties outside the platform entirely.

Final Thoughts

X’s 2026 privacy updates reflect a broader industry trend — platforms are collecting and using more data, particularly for AI training, while opt-out mechanisms remain technically available but not always easy to find. Pairing a manual review of your in-app privacy settings with a reliable, audited VPN gives you meaningfully better protection than relying on either approach alone. As with most digital privacy decisions, the goal isn’t perfect anonymity — it’s closing the easiest and most common gaps that most users leave open by default.

Have you gone through and adjusted your X privacy settings since the latest update, or are you still running on old defaults? Let us know what you found in the comments.