←
Rajat Gahlot6,084 posts
Rajat Gahlot avatar

Rajat Gahlot

@bigrkg

→ Building QuillShield: the AI security stack for onchain finance → Co-founder @QuillAudits_AI → 8y+ in web3 security · 1.5K+ protocol audits

◷ Joined October 2018▧ 462 Photos and videos
Rajat Gahlot avatar
Rajat Gahlot@bigrkg·Aug 18

This is genuinely concerning for the security of every DeFi protocol over the next few months. In July, the UK AISI reported that recent open‑weight models such as GLM‑5.2 and DeepSeek‑V4 Pro already match the performance of closed‑frontier models released only four to seven months earlier. That represents a significant narrowing of the six‑to‑ten‑month gap they had measured throughout most of 2025. A few months is not zero, but it is short compared with the typical time defenders need to adapt and deploy protections. Open‑weight models make the situation especially dangerous because attackers can run them privately, without supervision and with virtually no guardrails. For most of cybersecurity history, sophisticated attacks remained relatively scarce because the people capable of executing them were scarce. That scarcity acted as a quiet stabilizer for the ecosystem. AI is removing that constraint in two ways simultaneously. First, For actors who are already capable, AI raises throughput. Skilled actors such as the DPRK, elite North Korean hackers, and offensive security teams benefit from AI, which simply raises their throughput. Tasks that once consumed scarce expert hours can now be automated or run semi‑autonomously. A team that previously had to select one lead from ten candidates can suddenly explore a hundred. Second, for less sophisticated actors, AI lowers the barrier to entry. More of the attack chain becomes something you can supervise rather than master yourself. Someone who could never have written an exploit can now guide a model that can. The result is predictable: increased output from those already capable and a growing pool of newly capable attackers. DeFi will experience this pressure first and most intensely, in my opinion, as it is already occurring.

Preetam 🥷 avatar
⟳ Rajat Gahlot retweeted
Preetam 🥷@raopreetam_·10h

AI can write smart contracts fast, but it can't be the last line of defence. At @EFDevcon Mumbai, I representing @QuillAudits_AI shared with @develpgmbh on how teams should ship code in the Age of AI. My PoV: Attackers might have the upper hand right now, but defensive AI is catching up fast too. Full Interview Below 👇

8:03
Rajat Gahlot avatar
Rajat Gahlot@bigrkg·10h

worst part of a helpdesk incident is that the phish can arrive from the real support address, so checking the sender won't save you. in 2025 attackers put a phishing link in a trezor ticket subject and let trezor's own auto-reply send it. if you have an open etherscan ticket, treat any new link in it as hostile for now. x.com/etherscan/status/21229…

Security Notice: Reminder Regarding Etherscan Support Communications Earlier today, we detected attempts to gain unauthorized access to our customer support and ticketing system (Freshworks). We are currently investigating the situation and taking precautionary measures. In the meantime, we would like to remind all users to exercise caution when communicating with anyone claiming to represent Etherscan Support. - Never share your private keys, seed phrases, passwords, or other sensitive information. - Be cautious of unexpected links, attachments, or requests for sensitive information, even if they appear to come from our official support channels. As a general reminder, Etherscan Support will never ask for your private keys or recovery phrases. We are taking this matter seriously and will provide further updates if necessary. Thank you for your understanding and continued trust.

Rajat Gahlot avatar
Rajat Gahlot@bigrkg·12h

two weeks late, but these are my notes from the future threats table at @safe forward defence summit in mumbai: • AI attacks that start as prompt injection • infra attacks • AI-based opsec • AI swarms as attackers, and swarms as defence • agent opsec • quantum vs today's cryptography • cyber models the one i keep circling is agent opsec. agents hold keys, read inboxes and open PRs now, and most teams can't tell you which agent did what last tuesday.

Post attachment
Rajat Gahlot avatar
Rajat Gahlot@bigrkg·14h

Shadow automation of a particular task or role isn’t delivering the expected ROI, only pain. We need to build end-to-end automation for a role by spending weeks on architecture and domain knowledge, then let it run.

Polymarket Money avatar
⟳ Rajat Gahlot retweeted
Polymarket Money@PolymarketMoney·Nov 16

BREAKING: Nvidia unveils a quantum computing breakthrough that cuts a 12-week computing task to just 45 minutes.

Rajat Gahlot avatar
Rajat Gahlot@bigrkg·Nov 16

And this is when a 4-engine government in Delhi, and the same situation for air pollution.

River of toxic foam—Chhat Puja on the Yamuna, captured by @shashiskashyap. Saw this precise scene driving over Delhi’s Kalindi Kunj bridge yesterday. Imagine the generations of disease & cancers this will cause for these families. 💀

Post attachment
QuillAudits avatar
⟳ Rajat Gahlot retweeted
QuillAudits@QuillAudits_AI·Nov 14

Team Quill is attending Solana Breakpoint in London Our Head of BD & Sales, Akshadh, will be on the ground from Nov 15–17 at Olympia London. Looking forward to building new connections and talking through the current topics shaping Solana security, OpSec, RWAs, payments, and how teams are securing the Token Supercycle. If you’re building on Solana or heading to Breakpoint, come say hi. Happy to talk audits, QuillShield, or full lifecycle security. DM us or find @Akshadh1 at the event. See you in London.

Post attachment
Paras Chopra avatar
⟳ Rajat Gahlot retweeted
Paras Chopra@paraschopra·Nov 13

By the time Fable 6 comes along, perhaps Anthropic will simply prompt it to build all of SaaS automatically and bundle it in one Claude subscription. They will be the new Microsoft. Startups being created and run by AGIs is going to be the default going forward.

Rajat Gahlot avatar
Rajat Gahlot@bigrkg·Nov 13

So cool, much needed 🔥

Introducing Claude Meetings: add Claude to your team’s video call and it builds docs, whiteboards, and other artifacts with you while you’re still talking. Starting to roll out in beta on Team and Enterprise plans.

1:03
Rajat Gahlot avatar
Rajat Gahlot@bigrkg·Nov 12

Indra sent Menaka to Vishwamitra. Beijing allegedly sent Alice to Jensen. Vishwamitra lost his tapasya. Jensen lost Nvidia. Satyug had Menaka. The AI age has Alice Huang.

Post attachment
Rajat Gahlot avatar
Rajat Gahlot@bigrkg·Nov 12

Good orchestrator without brain/thinking. @bot is so dumb now a days. We must have the option to plug in with @claudeai Fable or Opus.

grok bot is the BEST AI agent orchestrator right now and THIS is the best setup i’ve EVER seen youtu.be/1cWZ7Ekff6E?si=YkcM… it runs claude code for engineering and paper for design here is the FULL setup and how to copy it:

48:31
Rajat Gahlot avatar
Rajat Gahlot@bigrkg·Nov 12

Very true, many things are draining the founders and the company overall.

PSA: creating wonderful products still requires extraordinary amounts of human focus, effort, motivation, and love. And an attention span.

Akshadh || Solana Breakpoint 🇬🇧 avatar
⟳ Rajat Gahlot retweeted
Akshadh || Solana Breakpoint 🇬🇧@Akshadh1·Nov 10

How are we, how are we! I'm heading to @solanaevents Breakpoint at Olympia, London, 15–17 Nov 🇬🇧 Around London for a few days, meeting people, catching up with familiar faces and seeing what's happening across the space. If you're around, give me a shout. Let's grab a pint! 🍻

0:02