Peter Bassill — latest writing on cyber security

Latest articles · page 13 of 13

• 0000 a5 2d 87 45 0004 fa 30 f7 68 0008 9e 9a 8d b2 000c 82 38 2a 9b 0010 ba 77 0d d3 0014 3c 6a 62 38 0018 b2 ba bd f4 001c c7 b1 09 8c 0020 08 17 e4 27
  Policy

  Digital privacy for board directors: the eighteen-post version

  An honest start to a long series. What digital privacy actually means for a board director in 2023, why the home / travel / work boundary is the right framing even though it leaks, and why children deserve four of the eighteen posts.

  • 7 February 2023
  • 6 min read
• 0000 b8 f6 c1 ac 0004 34 02 7e 18 0008 00 e3 9f 46 000c ea a2 46 ab 0010 e7 63 76 1f 0014 24 b2 8e 7f 0018 18 04 ee e1 001c 2d 25 b9 88 0020 ed 88 ef a6
  AI

  INT8 quantisation, in numbers — and why INT16 is the boring choice

  What "INT8-quantised inference" actually means once you do the arithmetic, why dropping from FP32 to INT8 is a cliff and dropping to INT16 isn't, and why every interesting question about putting an ML model on real silicon ends up here.

  • 26 May 2022
  • 11 min read
• 0000 3d 39 40 e2 0004 49 82 41 9c 0008 63 fe fc c2 000c 33 8a da 18 0010 ab 70 42 dd 0014 ca ca 32 fc 0018 83 c6 e4 09 001c 4f 31 64 56 0020 9e 6b 67 57
  Threats

  What the teenagers taught the Fortune 500

  LAPSUS$ compromised Microsoft, Okta, Nvidia, Samsung, Vodafone, and several others in a few months. They were teenagers using social engineering and MFA fatigue. The lesson, awkwardly, is that the dominant compromise vector in 2022 is social, not technical.

  • 26 April 2022
  • 8 min read
• 0000 54 ee 7a d8 0004 dc 5a 6a 4b 0008 2e 35 1f f6 000c 85 e0 20 06 0010 32 5b c3 68 0014 41 c3 9f f3 0018 ae 8d ab fe 001c 4e cd b8 d0 0020 ef c8 12 9c
  Threats

  Log4Shell, and the inventory question we cannot keep ducking

  A month on from CVE-2021-44228, the headline-grabbing exploits have slowed but the underlying problem has not. The discomfort of the past month was not really about Log4j. It was about how few firms could answer the question 'where is it running?'

  • 12 January 2022
  • 7 min read
• 0000 a1 e6 d9 fc 0004 8d aa ae b9 0008 91 f8 13 fd 000c b9 a9 3b 16 0010 b9 58 f0 26 0014 81 5a 0e 1e 0018 3d 3d 60 45 001c 2d 11 82 54 0020 9b c9 3e bd
  Threats / HbbTV: notes from the lab · part 4

  wlan0: the unlocked back door on every TV

  Part 4 of 4. Once you have root on the TV, the most useful thing on the device isn't the data on it — it's the second network interface nobody disabled. What this bypasses, why the SIEM is blind to it, and what to do about it.

  • 21 August 2021
  • 10 min read
• 0000 30 6f fc 31 0004 df 29 9b 13 0008 78 da a6 21 000c 7f 84 b4 ae 0010 9b 19 16 43 0014 97 49 c4 9d 0018 95 c7 05 f8 001c 6f 49 24 ec 0020 ac e5 13 a6
  Threats

  Pegasus, and the question for UK boards we have been pretending not to face

  The Pegasus Project disclosures last month confirmed what specialists have privately known for years: commercial spyware is a mature, well-funded industry, and its customer list includes governments most UK firms do business with. The board question is what to do about it.

  • 17 August 2021
  • 7 min read
• 0000 b2 98 68 18 0004 68 3e 20 83 0008 62 03 fb e6 000c f8 43 4f 15 0010 d6 80 c7 b2 0014 53 f2 d5 c6 0018 82 ab 56 1a 001c 58 a4 52 bb 0020 34 13 89 f1
  Threats / HbbTV: notes from the lab · part 3

  From the embedded browser to a shell on a smart TV

  Part 3 of 4. From the AIT-triggered page load to a shell prompt. CVE-2020-6383, shell.js, SMACK, and the public Samsung Q60T root chain.

  • 17 July 2021
  • 11 min read
• 0000 d7 be b4 9d 0004 ff bb 53 2e 0008 4c 48 74 b0 000c 1d a4 83 9f 0010 cc 87 8a a5 0014 0e ad e1 c2 0018 58 d9 9c 50 001c 98 bd 09 7e 0020 c2 cb b3 5a
  Threats / HbbTV: notes from the lab · part 2

  The lab rig: re-broadcasting HbbTV into a test bench

  Part 2 of 4. What I built on the bench to study HbbTV attacks safely. Hardware, software, the AIT injection step, and the legal bit (do not transmit DVB into open air).

  • 19 June 2021
  • 9 min read
• 0000 c2 3f 79 5b 0004 7b 70 23 3e 0008 ca a3 05 1c 000c 9a 14 6c 22 0010 0f 11 f7 39 0014 88 dc 6a 32 0018 18 fc a8 22 001c 74 39 6a 43 0020 f0 b6 a9 f2
  Threats

  Colonial Pipeline: the CNI lesson the UK should not need to learn the hard way

  Five weeks after the DarkSide ransomware attack on Colonial Pipeline shut down 45% of US East Coast fuel supply, what UK critical national infrastructure boards should be doing about it.

  • 10 June 2021
  • 7 min read
• 0000 d7 6e ab 50 0004 d2 b5 93 5d 0008 3f a8 54 34 000c cc 6a 61 c7 0010 1d e5 a2 9f 0014 18 ae d4 9c 0018 8a 0a ee 18 001c f4 a1 b0 dd 0020 ab 3a 33 be
  Threats / HbbTV: notes from the lab · part 1

  The TV in the corner: what HbbTV actually is

  Part 1 of 4. A primer on HbbTV from a security researcher's bench. Why I think the smart TV mounted on the meeting-room wall is the most under-considered attack surface in any UK office in 2021.

  • 22 May 2021
  • 7 min read
• 0000 dd 66 a6 78 0004 04 fb 6a 83 0008 28 a9 4a fc 000c 1b 3b 07 b2 0010 ea 80 9a ce 0014 c0 5a 68 cd 0018 0f d3 e1 9b 001c b5 cc a9 35 0020 8e 1e 3a 01
  Threats

  Hafnium and the patch-window asymmetry

  Five weeks after the Microsoft Exchange ProxyLogon disclosure, the dust is settling on what may turn out to be the most consequential mass-exploitation event of the decade. What it teaches us is structural, not tactical.

  • 6 April 2021
  • 7 min read